🌐 The Rotation Filter: Defense Cools, AI Hits the Power Wall & GLP-1's Efficacy War
The Ruck Filter #021 • June 1, 2026
Read time: 8 minutes
Welcome back to The Ruck Filter.
The S&P 500 just closed its ninth straight weekly gain - the longest winning run since 2023 - and the Dow crossed 51,000 for the first time. Two forces did the lifting: a reported 60-day extension of the US-Iran ceasefire that pulled Brent back to around $92, a one-month low, and a wave of AI-infrastructure earnings that revived the hardware trade.
But the index level hides the more interesting story. Underneath a calm tape, capital rotated hard this week - out of the obvious 2025 winners and into four places the headlines are still underpricing.
This week we filter the defense trade’s first real cooling, the shift of AI capital from compute into optics and memory, the power bottleneck now reshaping who supplies the data center, and the efficacy arms race quietly redrawing the GLP-1 winners.
A caution worth holding before we start: the Conference Board reported this week that CEO confidence fell back into negative territory in Q2 2026, reversing Q1’s optimism. Nine green weeks and deteriorating boardroom sentiment rarely coexist for long.
1. Signal vs. Noise: The Defense Trade Stops Going Straight Up 🛡️
The Noise: Ukraine ratified a €90 billion EU loan and Sweden is sending Gripens — defense stocks are off to the races again.
The Alpha: The pop on that news was real but shallow, and the more important signal this week was the sector starting to consolidate after a near-vertical 2025–26. For long-term capital, that distinction matters more than the headline.
The Filter: Three observations.
First, the immediate reaction was genuine. After Ukraine’s parliament ratified the €90 billion ($104.6 billion) EU loan agreement on May 28, European defense names jumped — Saab topped the Stoxx 600, up 7.4% on reports it will supply up to 150 Gripen fighters to Ukraine, with Renk +5.4%, Exail +13.2% and Rheinmetall +4.2%.
Second, by Friday the narrative had already shifted to consolidation. The same names that led the rally are now the subject of “is the boom cooling?” coverage — the sector has run so far that even good news produces shorter-lived moves. That is what a maturing trade looks like.
Third, and most important for sizing: Ukraine procurement is marginal to the actual thesis. Rheinmetall carries roughly €1.7 billion of Ukraine orders inside a backlog above €60 billion. The real driver remains structural — NATO’s push toward 3.5% of GDP, Germany’s €500 billion infrastructure fund, and the debt-brake exemption for defense spending that runs through 2029. Those do not turn on a single loan vote.
A measured approach:
Rheinmetall (RHM.DE): still the cleanest large-cap proxy for European rearmament, guiding to roughly 45% revenue growth in 2026. After the run, this is about averaging into weakness rather than chasing strength. What to watch: backlog conversion and margin on the next print - order intake is not the question, execution is.
Renk (R3NK.DE) and Hensoldt (HAG.DE): the higher-beta DACH plays (transmissions and sensor/optics). They amplify both the upside and the consolidation. Appropriate only as sized satellite positions.
Saab (SAAB-B.ST): the export-momentum name after the Gripen news, though much of that optimism is now priced. What to watch: whether the 150-jet letter of intent converts to a firm contract.
The discipline point: a sector that no longer rises on good news is telling you something. For patient capital, the better risk-reward is buying the next pullback, not Friday’s spike.
2. The AI Trade Moves Down the Stack: From Compute to Optics and Memory 🔌
The Noise: Nvidia already reported a record quarter - the AI hardware trade has nothing new to offer.
The Alpha: This week the money rotated past Nvidia into the connective tissue of the data center: optical interconnect and high-bandwidth memory. That is where the fresh re-rating happened.
The Filter: Three earnings events this week told one story.
First, Marvell (May 27) posted record Q1 revenue of $2.418 billion, +28% year-over-year, and guided Q2 to $2.7 billion, +35%. Management raised its revenue outlook for both fiscal 2027 and 2028 on what it called exceptional AI-related bookings. The strategic tell is the M&A: Marvell closed its acquisition of Celestial AI, a photonic-fabric interconnect company, alongside XConn. Optical interconnect - moving data between chips with light instead of copper - is becoming the bottleneck as AI clusters scale, and Marvell just bought its way to the front of it.
Second, Dell (May 28) lifted its fiscal 2027 revenue outlook to roughly $167 billion, including $60 billion of AI servers, up from a prior $140 billion guide and well above the $142 billion consensus. The stock had its biggest move since its 2018 return to public markets. AI demand is no longer a promise on Dell’s income statement - it is the income statement.
Third, Micron crossed a $1 trillion market capitalization this week, up roughly 88% in a month, on record quarterly sales near $23.9 billion. High-bandwidth memory for AI accelerators has turned a historically cyclical commodity business into a structurally tight one. The next earnings report on June 24 is the test of whether the re-rating holds.
The through-line: as compute itself becomes Nvidia’s domain, the incremental investment opportunity moves to what surrounds the GPU - the optics that connect them and the memory that feeds them.
A measured approach:
Marvell (MRVL): the most direct listed way to own AI optical interconnect and custom data-center silicon. The Celestial AI deal validates the photonics thesis that has been circulating for a year. What to watch: whether the raised FY28 guidance survives the next two quarters, and integration of two acquisitions at once.
Micron (MU): the HBM and DRAM proxy, now priced for the cycle staying tight. After an 88% month, this is a position to scale into carefully, not at the top. What to watch: the June 24 print and any sign of HBM pricing softening into 2027.
The European angle: the same logic flows to the test and materials layer we have flagged before - vertical MEMS probe cards and advanced packaging films. The optics theme also points toward smaller photonics names, though most carry venture-grade risk and belong in the satellite sleeve of a portfolio, not the core.
3. The Physical Layer Bottoms Out: The Power Bottleneck ⚡
The Noise: The constraint on AI is chips. Build more accelerators and the data centers follow.
The Alpha: The binding constraint has quietly moved from silicon to electricity. Grid interconnection queues now run years, and operators are bypassing the grid entirely - which is creating a fast-deploy boom for decentralized power, with the clearest beneficiaries listed in the DACH region.
The Filter: Section 2 traced AI capital from compute into optics and memory. It does not stop there. It bottoms out at power and copper — and two events this week made that concrete.
First, on May 27, Germany’s 2G Energy announced the largest order in its history: a North American data-center customer ordered containerized combined-heat-and-power systems in the low-triple-digit megawatt range, with deliveries starting in the second half of 2026 and running for years. The stock rose roughly 11% on the day to around €66. The company now guides 2026 revenue to the upper end of its range and projects 2027 sales up about 20% to €570–620 million at an EBIT margin above 11%. This is a Mittelstand CHP maker - over 10,000 systems installed worldwide - repositioning into AI power infrastructure in real time.
Second, the Advent-backed gas-engine maker Innio, whose Jenbacher units fast-start in under five minutes and run on hydrogen blends, launched its IPO roadshow this week, targeting a valuation reported as high as $15 billion. The numbers behind it explain the timing: Innio’s data-center equipment order intake reached $2.28 billion in 2025, and total equipment order intake rose 188% year-over-year, with backlog near $4.8 billion by the end of March.
The logic is the same in both cases. A hyperscaler cannot wait three to five years for a grid connection while its competitors deploy GPUs now. Behind-the-meter gas and CHP plants can be commissioned in months, off the grid’s timeline. Power has become the long-lead item, and whoever can deliver megawatts fast holds the pricing power.
A measured approach:
2G Energy (2GB.DE): the cleanest listed DACH proxy for fast-deploy decentralized power, now with a marquee US data-center order. Small-cap and lumpy: revenue moves with large orders, so expect volatility. What to watch: the Q1 update due June 2 and whether additional data-center orders convert. One honest governance flag — the company has delayed its preliminary full-year figures into mid-June due to an ERP transition, worth noting for a position of any size.
Innio (pending IPO): direct exposure to behind-the-meter gas engines, but discipline matters most at an IPO. The better approach is to let the first print settle and wait for the first quarter of public reporting rather than chasing the debut. Reported valuations near 27x forward EBITDA leave little room for disappointment.
Wärtsilä (WRT1V.HE): the larger, more liquid Finnish way to own gigawatt-scale flexible power for US data centers, with a deep service backlog that smooths the cyclicality.
Prysmian (PRY.MI): the world’s largest cable maker, a structural beneficiary of the same build-out, since every data center and offshore wind farm needs the copper and fiber to connect it. The lowest-volatility way to play the physical layer.
4. GLP-1’s Next Phase: The Efficacy War and the Differentiated Player 💊
The Noise: Eli Lilly and Novo Nordisk have settled the obesity market between them - the trade is done.
The Alpha: This week reframed the contest along two new lines: an escalating efficacy benchmark, and Lilly using its cash to diversify away from a one-drug story. Both reward different kinds of investor.
The Filter: Three developments.
First, the efficacy bar moved. Late-stage data for Lilly’s triple-acting drug retatrutide showed up to 28% weight loss - a new high-water mark that resets what “best in class” means. Lilly shares have risen roughly 25% in a month, the strongest among major obesity names.
Second, Lilly diversified. This week the company announced deals worth up to $3.8 billion to acquire three vaccine developers, pushing deeper into infectious disease. This matters for portfolio construction: it signals Lilly is building a second pillar rather than betting everything on the obesity franchise - exactly the kind of capital allocation long-term holders should want to see at this valuation.
Third, the contrarian case on Novo sharpened. Novo Nordisk is being priced as if it has lost the race, trading around 18x forward earnings with a high yield, despite its Wegovy pill posting the strongest GLP-1 launch on record - over two million US prescriptions since its January debut. CEO Mike Doustdar spent the week pushing the narrative beyond pure weight-loss percentages toward broader outcomes in cardiovascular, kidney and liver health, where the long-term value may actually sit. Separately, Truist initiated coverage of the smaller, differentiated player Viking Therapeutics with a buy rating and an $83 target.
A measured approach:
Eli Lilly (LLY): the momentum and pipeline leader, now diversifying its risk. Around 33x forward earnings prices in continued near-flawless execution; suitable as the quality anchor of a healthcare sleeve, less so as a fresh full position at these multiples. What to watch: retatrutide’s path to filing and the integration of three vaccine assets.
Novo Nordisk (NVO): the contrarian value side of the same structural story. Cheaper, higher-yielding, and arguably mispriced if the outcomes data lands. What to watch: whether the “health outcomes” narrative gains clinical traction before Lilly’s oral drug scales globally.
Viking Therapeutics (VKTX): the high-risk, high-reward optionality on a differentiated mechanism. A small position at most; this is binary on trial data, not a compounder.
The structural reminder: the consumer-side pressure on the “calorie industry” — packaged snacks, fast food — does not move on any single week’s news, but the efficacy gains compound the long-term case against it. A relative-weight call, not a short.
Outro: The Rotation Beneath the Calm
A flat-looking tape can hide a violent rotation. This week the index barely moved, but capital left the obvious 2025 trades - straight-line defense, pure-compute AI, single-drug pharma - and moved toward what comes next: the consolidation entry, the optics-and-memory layer, the power and copper beneath the data center, the diversified pipeline.
The discipline for long-duration capital is to resist two temptations at once: chasing the names that already tripled, and dismissing the rotation as noise. Nine green weeks with souring CEO sentiment is a market worth respecting, not crowding into.
The Takeaway: When the headline trade stops rising on good news, are you positioned in the rotation that follows - or still holding last year’s winner?
Daniel Ruck Editor, The Ruck Filter
Filter Sources this week
Bloomberg & CNBC: US and European market wraps, May 28–30, 2026
CNBC: “European defense stocks cooling off,” May 30, 2026
Marvell Technology Q1 FY27 results (SEC Form 8-K), May 27, 2026
Bloomberg: Dell FY27 AI-server outlook, May 28, 2026
2G Energy AG ad-hoc release: largest order in company history, May 27, 2026
INNIO Group: IPO roadshow launch & SEC Form S-1, May 2026
Stocktwits / company disclosures: Lilly vaccine acquisitions & retatrutide data, May 28, 2026
The Motley Fool: Novo Nordisk contrarian case, May 28, 2026
The Conference Board: Q2 2026 CEO Confidence
Disclaimer: The Ruck Filter is for informational purposes only and does not constitute financial, investment, or tax advice. The information provided is based on data available at the time of writing and is subject to change. Investing in financial markets involves risks, including the potential loss of principal. Every reader is solely responsible for their own trading and investment decisions. Please conduct your own due diligence or consult with a licensed professional before making any financial commitments.