Two months into the portfolio. We’re +17.5% since inception and in May/26 we went +9.3%. The MSCI World Index returned +12.5% since inception and +5.2% this month, meaning we’ve generated 500 bps of alpha in our first sixty days. For those who like to compare with the S&P 500, the index is +15.6% in the same period, so we’re ahead on 190bps as well. Once again, this is a signal that the framework is doing what it was built to do: a deliberate blend of growth and defensives, calibrated for high returns with controlled volatility.

This is our allocation right now and the total return since inception.

• Sharpe ratio: 1.53

• Annualized volatility: 20.3%

Three insights:

• MU is the absolute outlier: 84.5% annualized vol, yet it delivered +167% over the period. Largest risk contributor, but also the portfolio’s primary return driver. The small sizing was well-timed.

• The portfolio’s realized vol is surprisingly low (20.3%) given the high individual volatility of its components, diversification and imperfect correlations doing exactly what they’re supposed to.

• PG and ACGL are the only losers over the period: the two most defensive names in the book underperformed as the market rallied through April and May.

This month, the highlight was Micron (MU) that went +86%. When I first bought, it was a 4% allocation of my portfolio, now it’s already 10%, which makes me think I should probably trim, but the market continues to push forward and I’m still not seeing the end of this crazy rush.

This is the May/26 return of the portfolio.

As we can see, the returns were mainly driven by tech companies suchs as ASML, Micron and Microsoft. But, we also had a great result with IBKR, our main exposure and CVS, that is a turnaround case, that is going really well since we bought. The defensive end did what it should do, kept volatility lower and defended the portfolio.

Trade diary

This was the only trade we did this month and we’re proud of it. The idea is not to overtrade and we managed to keep our cool.

Closed INTU: thesis oficially broken

Closing the position and booking a loss after earnings.

• Moat cut from Wide to Narrow by Morningstar on AI risk and weaker switching costs.

• Q3 FY26: revenue miss, growth slowed to 10% YoY, TurboTax guide trimmed to 7%.

• 17% headcount cut plus $300–340M in restructuring screams structural, not cyclical.

• IRS Direct File now in 25 states, a real threat to TurboTax’s dominance.

• Analyst targets reset lower; stock looks “cheap” for the wrong reasons.

Initiating GOOG: a quality tech

Opening a small position in a great company. Better late than ever.

• Berkshire tripled Alphabet stake

• Dominant search monopoly

• AI growth potential

• More attractive valuation than Amazon

• P/E 34x and ROE 32%

• Low RSI (34)

Upcoming Events

One print inside the next four weeks deserve attention:

→ Micron earnings scheduled for June 24

We’ll size into and around this accordingly.

Onward to month 3.

The Capital & Cycles Equities Portfolio is a transparency exercise, not a recommendation. Positions and trades are disclosed for didactic purposes. Nothing here constitutes investment advice. If you found this useful, the best thing you can do is forward it to one friend who thinks about markets.

Subscribe now

Share

Leave a comment

A relatively orderly close to a genuinely turbulent month.

• The Middle East drifted further toward de-escalation

• Oil settled comfortably below the wartime peak

• The Fed held in line with expectations at its April meeting

• Hyperscaler earnings re-validated the AI capex cycle

  Capital & Cycles is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

  ---

Where we stand

One month into the portfolio. We’re +8.2% since inception. The MSCI World Index returned +7.3% over the same window, meaning we’ve generated roughly 90 bps of alpha in our first thirty days. We did not expect that gap to open this quickly, and we’re not going to get carried away with a one-month sample. But it’s a useful early signal that the framework is doing what it was built to do: a deliberate blend of growth and defensives, calibrated for high returns with controlled volatility.

• Sharpe ratio: 1.27

• Annualized volatility: 22.3%

The defensive design earned its keep this week. Just before the hyperscaler prints, the defensive bucket was up while the growth names were down. Four “boring” names offset roughly 61% of the drawdown coming from the other eight. In plain English: the defensive sleeve did its job on a day when the growth sleeve did not.

A book that only goes up when everything goes up isn’t a book, it’s a beta tracker. The point of construction is to win on net, across regimes.

Trade diary

I made more trades this month than I’d ideally want. Truth in advertising. But this was not a quiet month: a Middle East war, a software drawdown, an AI-bubble narrative crystallizing in real time, hyperscaler prints, the full Q1 earnings cycle and a Fed meeting, all inside thirty calendar days.

The two trades that closed out the month are below.

Closing ACN: the thesis broke faster than the stock

We initiated ACN as a small starter position in week three on a “quality compounder at a value multiple” thesis. Last Friday, we closed it. The trigger was the earnings release:

• FY26 revenue guide of $71.8–73.2B against $73.9B consensus, with a soft Q3 guide layered on top, not a one-quarter wobble, an outlook reset.

• Federal exposure now a ~1% revenue drag as DC budget priorities shift.

• Large enterprise IT projects continue to slip to the right, the consulting-to-implementation handoff is the most exposed link in the chain.

• The stock is already crushed: down 38% from its 52-week high, -31% YTD, printing fresh lows around $175. But it’s still trading on a “quality compounder” narrative that the print just dismantled.

When the operating story stops matching the multiple story, it’s the multiple that gives. Sitting through that gap, in our view, is paying tuition for a lesson the market has already taught.

We exited cleanly. Cutting the slow-bleed and upgrading future optionality.

Initiating ACGL: boring insurance, spicy numbers

The redeployment went into Arch Capital Group, a 4% starter position.

ACGL is, in our view, the only high-quality compounder still trading at a single-digit multiple in its peer set. The fundamentals, frankly, don’t look like an 8x stock:

• P/E 8.3x with ROE 21%

• Zero net debt

• 8 consecutive earnings beats (12–23% surprise per quarter, that’s not luck, that’s a model)

• Combined ratio 80.6%, exceptional underwriting discipline

• $26.9B capital base generating elevated investment income in a higher-for-longer regime

• GQG Partners doubled their stake to 1.7M shares in Q4 2025; Bridgewater added, quality institutional money is layering in

• Consensus target $109–112 vs. current ~$94, implying 12–15% upside

Compare that to Chubb at ~13x. The peer-multiple gap deserves to compress.

The macro setup is also well-aligned: reinsurance rates softening is a real headwind, but ACGL’s posture is to reduce volume rather than chase unprofitable business. That underwriting discipline is exactly the management quality we want compounding our exposure.

The position also diversifies the book away from software/IT services at a moment when that sleeve is under pressure, a side benefit that meaningfully tightens our risk profile.

Boring insurance. Spicy numbers. 4% starter, room to add.

Earnings recap: the names we hold reported

Q1 print season landed inside our window for five holdings. The run-through:

Procter & Gamble (PG)

A clean print that justifies the boring weight in the book.

• Net sales $21.2B, +7.4% YoY (vs. $19.8B prior)

• Net earnings $3.93B; EPS $1.63 (up from $1.54)

• Operating income flat at $4.58B, higher sales offset by SG&A and product cost pressure

• Restructuring in motion: $782M of charges over nine months, anchoring a 7,000-position headcount reduction designed to widen future operating leverage

• Total assets $128.4B, $3.17/share in common dividends returned over the nine-month period

Translation: top-line growing, margins temporarily compressed and management is actively re-engineering the cost base. PG is doing exactly what a defensive should do, compounding quietly through a noisy macro.

Outlook: cautiously optimistic.

Visa (V)

The blast print of the month.

• Net revenue $11.23B, +17% YoY (vs. $9.59B)

• Net income $6.02B; EPS $3.14 (up from $4.58B prior-year net income)

• Data processing revenue +18%

• Value-added services revenue +27%, totaling $3.3B for the quarter

• Closed $1.5B acquisition of Argentine processors Prisma and Newpay, deepening the biometric/tokenization stack

• Reduced debt by $4.0B in H1

• $11.6B in buybacks executed alongside

Visa is doing the rare quadruple: accelerating organically, compounding via M&A, deleveraging and returning capital, all in the same quarter. The stock printed +8.8% the next session. When a $600B+ name moves like a mid-cap on a single print, it’s because the print genuinely changed the model.

Outlook: bullish.

EMCOR Group (EME)

The non-obvious AI play continues to deliver.

• Revenue $4.63B (vs. $3.87B prior — a +19.6% YoY print)

• Net income $305.5M; EPS $6.84 (up from $5.26)

• Total remaining performance obligations of $15.62B, with $12.17B expected to convert within twelve months

• Cash $916.4M (down from $1.11B, reflecting $87.1M of buybacks and normal working-capital fluctuation)

• Debt-free balance sheet (ex finance leases)

• Quarterly dividend raised from $0.25 to $0.40, a 60% hike, which is a serious capital-allocation signal

The data center buildout is showing up in EMCOR’s order book in real time. The dividend hike says management has the visibility to commit beyond a single quarter.

Outlook: bullish.

Microsoft (MSFT)

The print that anchored the entire week.

• Revenue $82.89B (vs. $70.07B prior; vs. $81.4B expected — clean beat)

• Net income $31.78B; EPS $4.27 (vs. $4.06 expected — ~5.2% surprise)

• Eight consecutive EPS beats — execution remains unusually reliable

• Microsoft Cloud at $54.5B, +29% YoY — core thesis intact

• Azure +40% YoY; AI revenue run rate above $37B — demand is real

• Net income +23% YoY with operating margin still elite

• Debt-to-equity 0.10, current ratio 1.39 — fortress balance sheet

• Quarterly capex $30.9B — the single point of bear ammunition; the AI buildout is expensive, and the eventual ROI debate isn’t going away

• OpenAI dilution gain of $7.25B in Other Income (9-month period) following the October 2025 recapitalization

• Consensus Buy (64 buys / 16 holds), average target ~$572.76

Capex is the only place a critic can plant a flag. Everything else compounds the bull case. We added to MSFT in week four; we are glad we did, despite the negative return so far.

Outlook: bullish.

Arch Capital (ACGL)

The print that landed immediately after we initiated.

• Reported EPS $2.50 vs. ~$2.47 expected — still a beat on the headline

• $87M net realized investment loss — the soft spot of the quarter

• Cat-loss ratio impact noted, with the core combined ratio ex-cat holding steady

• Stock fell 4.04% to $93.75 on April 29

This was a “good headline EPS, weaker composition underneath” setup, not an expectations failure, but the mark-to-market on the investment book disappointed on optics. We’re not framing it as a thesis-breaker. The underwriting discipline and capital-base story are exactly what we bought, and neither was impaired by this print. If the market keeps marking the name down on composition optics, we’ll add at lower prices, not exit.

Outlook: unchanged.

Upcoming events

Three prints inside the next four weeks deserve attention:

• CVS Health — Q1 earnings, May 6 (before market open)

• Intuit — Q3 earnings, May 21 (after market close)

• NVIDIA — Q1 earnings, May 27 (5:00 PM ET)

We’ll size into and around these accordingly.

Subscribe now

Three lessons from month one

A few honest reflections, because a transparency exercise that only highlights the wins isn’t actually transparent.

One — the defensive sleeve earned its keep. When the book started, I worried it would mostly drag the P&L in a risk-on tape. It did, on certain days. But on the days that mattered most, the hyperscaler-print volatility, the Iran headlines, the defensives absorbed losses we couldn’t have absorbed otherwise. Process over outcome, but also, the process worked.

Two — I overtraded. Macro chaos is a poor excuse for activity. A few of this month’s smaller moves probably netted to zero alpha after frictions. Next month I’ll aim for half the trade count and twice the conviction per trade.

Three — earnings discipline matters more than entry timing. ACN and ACGL both demonstrate this in opposite directions. ACN we exited because the thesis broke. ACGL we’re holding through a noisy print because the thesis didn’t. The framework is the same; the outcome is what differs. Position management starts with knowing what would force you out and then honoring it.

Onward to month 2.

The Capital & Cycles Equities Portfolio is a transparency exercise, not a recommendation. Positions and trades are disclosed for didactic purposes. Nothing here constitutes investment advice. If you found this useful, the best thing you can do is forward it to one friend who thinks about markets.

Subscribe now

Share

Leave a comment

Twenty-four sessions into the portfolio. We’re +6.8%.

Last week gave us big moves on fast-traveling headlines. This week gave us digestion. The reality of running an equities portfolio is that most weeks look like this: the names you sized correctly do their job, the names you didn’t get your attention and the tape rewards the discipline of not overtrading into noise.

We made one transaction this week. That’s it. And it was less about adding a new thesis than about managing an existing one with respect for what had already been earned.

The macro backdrop: a tape in digestion

The post-Hormuz rally finally took a breath. With oil stabilized in about $100 and the geopolitical premium quietly bled out of the tape, the market’s attention shifted back to where it belongs this time of year, Q1 2026 earnings. What we saw was a market that’s become, for now, a bottom-up market: names trading on their own fundamentals rather than on macro risk-on/risk-off switches.

That’s a healthier regime for a stock-picker’s book than a headline-driven one. It also means gains will come slower and more selectively from here and the edge sits in positioning, not prediction. The rally isn’t broken. It’s consolidating.

The only trade: trimming IBKR ahead of the print

On April 20, we sold a littlle bit of IBKR and redeployed the proceeds into MSFT. A virtually 1-for-1 cash rotation inside the book.

The rationale deserves some care, because this one runs counter to last week’s messaging, where we’d actually added to IBKR ahead of the print, on the back of read-throughs from JPM and Morgan Stanley’s trading desks. So why trim now?

Because the setup changed in our favor and when the setup changes, the trade changes.

Between our add and this trim, IBKR had run hard, appreciably so. The kind of move that takes a normal pre-earnings long into a crowded pre-earnings long. Position sizing and P&L are one thing; positioning risk is another. The question we asked ourselves wasn’t “do we still like IBKR for the earnings read-through?”, we do. The question was: at this price, into this print, what’s the asymmetry?

That’s the discipline. We were paid for taking the risk on the way up. We don’t owe the market an encore. When the expected value changes, the holding changes, even when the thesis is intact.

One important nuance: we didn’t exit IBKR. We trimmed. The name is still our largest position at 15% of the book. The long-term thesis, a structurally advantaged brokerage with best-in-class operating leverage to volume and rate cycles, remains our single highest-conviction expression in the portfolio. This was a size-down into strength, not a loss of faith.

Why MSFT as the re-deployment: same logic as last week, now compounded. MSFT is the quality ballast that lets us stay aggressive elsewhere in the book. Revenue growth 14.9% YoY, operating margin 46.7%, fortress balance sheet. Adding into it after a week of risk-on behavior is our way of raising the quality floor of the portfolio without going defensive.

What we did not do and why that matters

The harder call this week wasn’t the trim. It was leaving the other eleven names alone.

A few were genuinely tempting:

• NVDA continued its grind higher. The reflex is to take some off the top.

• ASML recovered meaningfully from the sell-the-news reaction to its guide. The reflex is to claim victory.

• MDLZ and PG have lagged for two straight weeks. The reflex is to cut the laggards.

We sat on our hands for all three. Here’s why:

On NVDA: our work suggests the AI capex cycle has a multi-year runway and we’re not in the business of trading the single most important name in that cycle on a weekly basis. Ride it until the thesis breaks.

On ASML: the guide disappointment last week was exactly the kind of noise our thesis is built to absorb. Selling on the recovery would mean we fundamentally didn’t believe what we were buying. Monopoly assets deserve patient capital.

On MDLZ/PG: they’re supposed to lag in a risk-on tape. That’s the whole point of carrying them. If we cut staples every time they underperform in a rally, we’re not running a diversified book, we’re running a growth-chasing one. And we already know where that ends.

The meta-point: in weeks like this, the alpha isn’t in the trade you make. It’s in the trades you don’t make. Turnover for its own sake is a tax, not a strategy.

Reading the tape from here

A few things we’re watching into next week:

1. IBKR earnings. We trimmed into strength. If it disappoints, we have capacity to re-add into weakness and the trim was a risk-management move, not a thesis change. If it prints clean, we left some P&L on the table; we’ll live with that.

2. Software and the NOW print that complicated the story. ServiceNow reported this week, and the read was ugly enough to matter. Guidance softened, the commentary around enterprise deal cycles wobbled and the tape punished the entire software complex in sympathy. That’s a meaningful data point and it cuts against the bullish Adobe thesis we were circling last week.

We’d been building a narrative that the AI-disruption fear in SaaS was overstated relative to fundamentals, and that the first software print to show resilience would force the sector to reprice higher. NOW was arguably the most important candidate to deliver that proof point. It didn’t. Whether the issue is company-specific (enterprise IT spend digestion, deal slippage) or sector-wide (AI genuinely compressing the growth runway) matters enormously, but we now have to wait for more prints to disambiguate.

Practical implication for our book: the ADBE trigger has moved further out. We’re not buying it this week and we’re not buying it next week unless we get a clean enterprise software print that rebuilds the thesis. “Cheap” is not a trigger. “Cheap plus a catalyst” is. On INTU, the Claude Mythos overhang still hasn’t manifested in observable business degradation, but after NOW, we’re more cautious about declaring victory there too. The Mythos bear case is now one disappointing print away from getting louder, not quieter.

3. Defensive sleeve re-rating. At 23% of book, we’re closer to the low end of the range where we feel comfortable running the defensive bucket. If the risk-on regime extends, we’ll eventually have to choose between rebalancing the weights back up or accepting a higher-beta book than we’ve been running. That’s a decision for week five or six, not this week.

Upcoming Events

• PG Q3 Earnings: April 24, 2026, Pre-market

• MSFT: Q3 Earnings: April 29, 2026, After-market

• EME: Q1 Earnings: April 29, 2026, Pre-market

Onward to week five.

The Capital & Cycles Equities Portfolio is a weekly transparency exercise, not a recommendation. Positions and trades are disclosed for didactic purposes. Nothing here constitutes investment advice. If you found this useful, the best thing you can do is forward it to one friend who thinks about markets.

Subscribe now

Share

Leave a comment

Seventeen trading days in, the portfolio is up 7.5%. That’s ahead of what we’d hoped for at this stage, but the gains have come from the right places, quality names compounding, tactical trims executed into strength and a rotation into growth that caught the tape at the right moment. Process over outcome, always. But we’ll take the outcome.

This week was one of those that separates a book from a watchlist. Real decisions, made in real time, with imperfect information. Here’s how it played out.

The macro backdrop

The week opened with the U.S. installing a blockade in the Strait of Hormuz, a tactical mirror of Iran’s earlier move. Call it what it is: a proportional, non-kinetic response. Not an escalation, an equalization. Markets priced it with remarkable composure. Oil ripped on the headline, but faded faster than most expected, never sustaining above the triple-digit handle. Equities traded mixed, with risk appetite visibly hesitant, but not capitulating.

Our read: if the tape is already fading the war premium, the incremental buyer wants to be long risk, not hedged. We chose to wait out the first 48 hours before touching the book. That patience paid, oil broke sub-$100 mid-week and equities went vertical. The complacency trade became the consensus trade and we were already positioned for it.

Software trade

SaaS has been in the penalty box for months. The market is pricing an existential AI disruption into every name with a monthly subscription. Our portfolio holds INTU and we took the opportunity last week to trim, but not cut.

The fear on the tape was that Anthropic’s latest model release, Claude Mythos, would accelerate the hollowing-out of vertical software businesses. But INTU isn’t a victim here, it’s a partner. That’s a category mistake the market keeps making.

Fundamentally, INTU is exactly the kind of name you don’t want to be underweight in a dislocation:

• P/E 23x on 16% revenue growth and 31% EPS growth

• 8 consecutive earnings beats

• Quality compounder trading at a value multiple

So why trim at all? Two reasons. First, the position had become sized-up relative to our risk framework, costly to hold through a binary print. Second, despite the eight-beat streak, we’re not willing to underwrite a ninth blindly. Trimming harvests some of the run, keeps the thesis intact and gives us dry powder if the stock gets handed to us post-earnings. It came back strong this week, exactly the sort of tape that rewards the discipline of not selling your winners.

ADBE is the more interesting name on the watchlist. The fundamentals are, frankly, absurd for the multiple:

• P/E ~13x with FCF yield 9–11%

• 89% gross margins, 37% operating margins, ROE 63%

• Revenue growing 12%, EPS growing 35%

• Eight consecutive beats

• Consensus target $350 — roughly 50% upside from here

But given we just trimmed INTU on the Mythos overhang, buying ADBE in the same week would be a mixed message to our own thesis. Risk management first. If the software sector reprices lower, we have a ready-made shopping list.

A broader point worth surfacing: credible reports suggest only about half of announced 2026 data centers will actually be built, permitting, power and grid interconnect are the bottlenecks. If that’s the base case, the market is dramatically mispricing a software apocalypse that simply can’t arrive on the feared timeline. The compute shortage is a feature for SaaS incumbents, not a bug.

One honest question for readers: do you personally know anyone who has cancelled a software subscription because of Claude? Neither do we. Drop your answer in the comments, this is the datapoint the sell-side isn’t collecting.

Defensives and the MSFT rotation

We added PG to the book last week. It lagged the growth names this week and that’s not a bug, it’s the design. A defensive consumer staples holding should lag in a risk-on rally. If it didn’t, it wouldn’t be doing its job as a portfolio beta-dampener. Fundamentals are clean, but let’s not romanticize it: 23x earnings on single-digit revenue growth is a high price to pay for calm.

Ray Dalio’s observation applies here, it’s very hard to beat the S&P if you diversify. True. But the goal of this book is not to beat the S&P by 400 bps with 30% vol. The goal is compounded, risk-adjusted returns with uncorrelated exposures. That’s what lets you stay in the seat.

That said, the tape was giving us a gift and we weren’t going to sit through a melt-up with oversized defensives. We trimmed 20% each of PG and MDLZ and rotated the proceeds into MSFT.

Why MSFT? There isn’t really a better single-name expression of “steady tech” on the market:

• Revenue growth 14.9% YoY

• Net income growth 15.5% YoY

• Gross margin 68.6%, operating margin 46.7%

• Debt-to-equity 0.10, a genuinely pristine balance sheet

• Cloud + enterprise software + AI monetization, all underwritten by real cash flows

• Consensus Buy, average target $582.56

This is diversified tech exposure without the white-knuckle volatility of a pure-play AI name. We’d rather own the infrastructure than the lottery ticket.

Financials week

Tuesday gave us JPMorgan earnings. Wednesday gave us Morgan Stanley. Both prints were critical to us, not because we own either (we don’t), but because our largest position is IBKR and these two releases were the cleanest possible read-through.

The signal was unambiguous: trading activity spiked materially, driven by the volatility cluster around the Iran/Hormuz situation. If that read-through holds for IBKR, which is structurally more leveraged to retail and prop trading volumes than the bulge brackets, their print should be excellent.

So we made the call. Right after JPM’s release on Tuesday, with oil already signaling the market was pricing out the war premium, we trimmed 50% of EPD (a more defensive midstream exposure) and added to IBKR ahead of earnings. Pure tactical allocation, high-conviction, short-duration, with a pre-committed plan to trim back into the print regardless of direction. That’s the discipline: size up the edge, take it off when the edge is gone.

Semis

ASML reported Wednesday. Q1 2026 beat on both profit and sales, but the near-term guide disappointed. Q2 2026 net sales guidance of €8.4B–€9.0B put the midpoint below consensus and the tape punished the name despite continued AI demand signals in the commentary.

Our take: this is a textbook sell-the-news, not a thesis-breaker. ASML is a monopoly in the most strategically important manufacturing step in the world economy. You don’t trade around that holding on a single guide; you own it through the cycle. Numbers remain exceptional, order book is intact, and the AI capex runway is multi-year.

TSM reported Thursday and delivered. Positive read-across for the entire semiconductor complex and a useful confirmation that the AI capex cycle is still very much alive at the foundry level. Good news for ASML holders willing to sit through noise.

Exiting EPD, initiating ACN

We closed out EPD this week. The thesis carried us well through the start of the run, but with the war premium deflating and the name having already had its spike earlier in the year, the risk/reward tightened. EPD is never a pure oil beta, but sentiment linkage is real and the setup no longer justified the weight.

We redeployed into ACN (Accenture), a name we’ve been circling for weeks. The setup is genuinely interesting:

• Down roughly 29% YTD and 33% over one year

• Still growing revenue 7.4% YoY, EPS 6.2% YoY

• ROE 23.7%, effectively no net leverage concern

• Asset-light, cash-generative, high-return — the model textbook-defenders actually like

• Consensus Buy, average target near $300 vs. current price near $195

• Trading cheaper than its premium consulting/IT peer set

The short version: the market is treating ACN like a broken growth name, but the business is still compounding, the balance sheet is fortress and the multiple no longer asks you to believe much. We started small — 5%. Room to add if the thesis plays out.

Portfolio state

We’re now at 12 names. Our hard cap is 15, so we have capacity to add conviction without bloating the book. The defensive share sits at 39%, meaningfully lower than last week, deliberately so. We are positioned for the continuation of this rally, with enough ballast to stay in the seat if it reverses.

The setup from here:

• If the war premium stays priced out and oil holds below $100, the risk-on rotation has more room.

• If software reprices lower on Mythos fears, ADBE becomes the next add.

• IBKR earnings are the nearest catalyst; we’ve front-run it with a plan to de-risk on the print.

  ---

Upcoming Events

•IBKR Q1 Earnings: April 21, Post-Market

•PG Q3 Earnings: April 23, Pre-Market

•EME Q1 Earnings: April 23, Pre-Market

•MSFT Q3 Earnings: April 29, Post-Market

Onward to week four.

The Capital & Cycles Equities Portfolio is a weekly transparency exercise, not a recommendation. Positions are disclosed for didactic purposes. Nothing here constitutes investment advice. If you found this useful, the best thing you can do is forward it to one friend who thinks about markets.

Subscribe now

Share

Leave a comment

This portfolio was built for a stagflation scenario, the base case we’ve held since inception. Sticky inflation, slowing growth and geopolitical disruptions feeding into supply-side pressure. This week provided a live stress test of exactly that thesis.

Going into April 7th, we had a hard rule: no new capital deployed until the Iran situation clarified. Trump’s threat to “end Iran’s civilization” over the Strait of Hormuz was a binary risk event with global consequences, the kind of thing that doesn’t price in gradually. Markets were in wait-and-see mode and so were we.

Sometimes the best trade is not touching anything.

That held. Trump announced a 14-day ceasefire and risk assets rallied hard. The move felt stretched almost immediately, the market traded as if the ceasefire were permanent and as if CPI were about to come in soft. Neither assumption was well-supported. Hormuz traffic hadn’t recovered, core inflation risk was still present and within days the ceasefire was already fraying: Israel continued operations in Lebanon, passage through the strait was running below 10% of pre-war levels and the so-called 10-point framework was being violated by multiple parties.

The rally gave us a window. We used it.

What we did

Trimmed — AI concentration above threshold

By April 8th, the post-ceasefire appreciation had pushed our combined AI/semiconductor exposure above 30%, a concentration level we’re not comfortable holding in the current environment. We trimmed three positions without fully exiting any of them.

NVDA - Nvidia

Still our highest-conviction AI name. Reduced purely on concentration, not on thesis. The CUDA moat and data center demand cycle are intact.

ASML - ASML Holding

New U.S. export restrictions targeting equipment sales to China added a layer of risk. We reduced exposure but remain invested, the EUV monopoly and order backlog visibility don’t disappear with one policy move.

MU - Micron Technology

Same China export restriction headline hit Micron directly. HBM thesis remains intact, but size reduction was warranted given the regulatory overhang and our concentration limits.

Trimmed — Thesis weakened

INTU - Intuit

Anthropic released a major Claude model update this week, which renewed pressure on AI-exposed software stocks. For Intuit specifically, the concern is real: if AI continues to commoditize tax and accounting workflows, the switching-cost moat we underwrote gets eroded at the margin. A -14% move in six days is too much to hold the full position without reassessing. We still like the company, QuickBooks and TurboTax are not going away tomorrow, but the sizing no longer reflects our conviction level relative to the risk.

Exited — Risk/reward no longer compelling

KR - The Kroger Co

We exited Kroger entirely. This is not a call on the company being in trouble, it's a capital allocation decision. At $70.28, the stock is trading just below consensus target of $74.25, which leaves less than 6% of implied upside. That alone would be manageable, but the valuation has stretched: a P/E of 45.5x is not cheap for a grocery business running 1.3% operating margins and profit growth has declined sharply year over year. The original thesis was built on a defensive setup with an embedded data monetization story. That story is intact, but the risk/reward has compressed relative to everything else available in the portfolio. When a position stops being asymmetric, capital has better places to be.

Added — Defensive rebalancing

Capital freed from the trims went into the three positions that provide the most meaningful diversification away from semiconductors and software. Negative correlation with tech is exactly what you want when volatility is the dominant regime.

CVS - CVS Health

Good timing: the government announced better-than-expected Medicare Advantage payment rates this week, relieving direct pressure on health insurer margins and removing a significant regulatory overhang. The thesis on CVS was always that the market was pricing distress into a business serving 100M+ Americans in essential healthcare categories. That view got reinforced.

EPD - Enterprise Products Partners

If the ceasefire breaks down, energy infrastructure wins. If it holds, EPD’s 7%+ distribution yield still pays you while you wait. The asymmetry is attractive in either scenario. Hard-asset cash flows belong in a stagflation portfolio.

MDLZ - Mondelez International

Consumer staples with global brand distribution and demonstrated pricing power. In a hot inflation scenario, MDLZ passes costs to consumers. In a risk-off scenario, people still buy Oreos. The defensive profile is exactly what this rebalancing called for. Besides, cocoa is in its lowest level since 2023.

New Positions

V - Visa Inc.

We exchanged cyclical software exposure for a sturdier large-cap franchise. Visa is one of the highest-quality businesses in global payments. 59% operating margin, 54% ROE, +15% YoY revenue in Q1 and a 5-year revenue CAGR of 13%. The stock is -11% YTD, weighed down by litigation risk and regulatory noise around interchange fees. We view that discount as an opportunity. The payment network itself, with billions of cards and decades of merchant acceptance, is not disrupted by a regulatory decision. Institutional positioning remains constructive. Implied upside relative to consensus is 30%. This is a world-class compounder on sale.

PG - Procter & Gamble

KR’s exit freed capital that needed a home in consumer staples, the defensive sleeve of the portfolio still needed coverage. PG is the right replacement. This is not a growth bet, it’s a durability bet. Tide, Pampers, Gillette, Oral-B, category leaders across everyday-use, repeat-purchase products sold in virtually every market on the planet. The kind of business that doesn’t need a favorable macro to perform, it just needs the world to keep washing clothes and brushing teeth.

The numbers support the entry: 22% operating margin, 17.8% profit margin, modest positive EPS growth and analyst consensus near $168 against a current price of $147, roughly 15% implied upside before any dividend contribution. PG trades at a premium to the market, and it deserves to. What we’re buying here is predictability in an environment where predictability is scarce.

Where the portfolio stands

The portfolio structure was always built for exactly this kind of week: EPD and the staples cover a hot CPI or a failed ceasefire; the growth names cover a soft CPI and a deal that holds. You don’t need to predict which one plays out when both scenarios have coverage. That’s the point of the construction.

Looking ahead

We are not convinced the ceasefire holds. Hormuz traffic remains well below pre-war levels and the geopolitical dynamics in the region do not support a clean resolution in 14 days. That makes CPI, already a volatility event, even more consequential. PCE came in as expected, but that data predates the conflict, so it offers limited signal for what’s next.

The portfolio is positioned to absorb both outcomes. That said, two earnings events next week will move our positions in ways that macro won’t:

April 13 — IBKR Q1 2026 earnings — largest portfolio position

April 15 — ASML Q1 2026 earnings — key read on AI capex and China exposure

Both reports matter beyond the single-stock reaction. IBKR’s earnings will speak to financial market activity and margin trends; ASML’s guidance will be one of the clearest real-time signals on where AI infrastructure spending is heading.

Expect more volatility than usual next week.

See you next week.

Disclaimer: This publication is for informational purposes only and does not constitute investment advice, a solicitation, or a recommendation to buy or sell any security. The positions and opinions expressed here reflect the author's personal portfolio and views at the time of writing. Past performance is not indicative of future results. Investing involves risk, including the possible loss of principal. Always do your own research and consult a qualified financial advisor before making any investment decisions.

Leave a comment

Share

Subscribe now

The premise of this publication is simple: document the thinking, show the work, and track the results in real time. No predictions dressed up as analysis. No hindsight. If the thesis changes, you’ll read about it here.

The plan is to post weekly, a consistent snapshot of where the portfolio stands, what moved and whether the underlying thesis on each position is holding. That said, trades can and will happen any day when the setup warrants it. When they do, you'll read the reasoning here after execution.

Enjoy.

The Framework

Every position in this portfolio starts with a macro question: where are we in the cycle and which sectors are structurally favored at this point? That filters the universe before a single stock is analyzed.

The current macro read: we are in a late-expansion phase with above-trend inflation expectations, tightening credit conditions at the margin and a structural AI investment super-cycle running in parallel. That creates a specific set of winners, businesses with pricing power, hard-asset cash flows, financial infrastructure that benefits from market activity and the pick-and-shovel layer of the AI buildout.

Once the macro filter is applied, individual stock selection follows three criteria:

1. Durable competitive advantage: The business must have a moat that is genuinely difficult to replicate, network effects, regulatory positioning, high switching costs, capital-intensive infrastructure or brand scale.

2. Capital allocation quality: We want management teams that have demonstrated consistent ability to allocate retained earnings at above-average returns. Return on invested capital over five years matters more than the last quarterly beat.

3. Reasonable entry point relative to intrinsic value: Growth is not a substitute for price discipline. Every position has an explicit entry thesis, a reason why the market is mispricing the asset at the time of purchase.

AI plays a specific role in the process: it is a research and stress-testing tool, not a signal generator. We use it to challenge assumptions, synthesize large volumes of earnings transcripts and filings and identify blind spots in our own reasoning. The final decision is always human judgment applied to the framework above.

We don’t predict cycles. We position for them and then let the fundamentals carry the weight.

The Portfolio

Ten positions. One account. The portfolio is concentrated by design, there are no diversification-for-its-own-sake positions here.

The Theses

Here is why each position is in the portfolio.

IBKR - Interactive Brokers Group

IBKR is the largest position because it is one of the few financial businesses that structurally improves as markets grow. The model is built on low costs, automated risk management and has global client base. ROE is about 21.3%, net income growth is about 54.9% YoY and the stock is up about 56.9% over the last year. Analysts still lean positive with a consensus Buy and average target around $84 versus a current price near $67.8. Interest income on client cash balances is a meaningful earnings driver, one that compounds quietly regardless of market direction.

Thomas Peterffy built IBKR over five decades. He owns a majority stake. That alignment matters. At our average entry, we were buying this compounder before the broader market fully re-rated it.

NVDA - Nvidia Corporation

Nvidia is one of the strongest businesses in the market: revenue growth is about 65.5% YoY, profit growth is about 64.7%, gross margin is 71.1% and ROE is an exceptional 76.3%. The AI infrastructure thesis is not hype, it is capital expenditure commitments. The four largest hyperscalers have guided for combined capex north of $300B in 2025-2026 and Nvidia captures a disproportionate share of that through GPU dominance.

Analysts are still constructive too: consensus is Buy with an average target around 278.6 versus a current price near 177.1.

ASML - ASML Holding N.V.

ASML give us exposure to the semiconductor buildout and it’s a monopoly. There is no other company in the world capable of producing EUV (extreme ultraviolet) lithography machines, the equipment required to manufacture leading-edge semiconductors at 5nm and below. Every chip at that node, from Apple’s M-series to TSMC’s most advanced logic, requires ASML hardware.

Analysts still rate it Buy with an average target of 1458.5 vs current price around 1327 and the stock is up about 16.8% YTD and about 103.8% over 1 year.

For our portolio, it’s a diversification inside semis.

The main risk? Earnings are coming on April 15.

KR - The Kroger Co

Kroger is a defensive anchor in the portfolio, grocery retail is one of the most recession-resistant categories that exists. People eat regardless of the macro cycle.

The underappreciated asset is Kroger Precision Marketing: a first-party data advertising platform built on 60M+ loyalty card members. That business generates high-margin revenue that does not show up cleanly in traditional grocery metrics and the market tends to value Kroger purely as a low-margin retailer.

Right now the setup is solid: the stock is around 72.4, up about 12.0% YTD, and analysts still lean Buy with an average target near 74.3 and a high target of 82. It’s not the most exciting stock on our portolio, but earns its place by lowering the volatility and balancing other names.

MDLZ - Mondelez International

Mondelez is also a defensive global stock for the portfolio. They own the snacking shelf in 160 countries: Oreo, Cadbury, Toblerone, Ritz, belVita. These are not just brands, they are shelf-space products with decades of distribution infrastructure behind them. Emerging market exposure (roughly half of revenue) provides a long organic growth runway as middle-class consumption rises across Asia, LatAm, and EMEA.

MDLZ demonstrated meaningful pricing power through the 2021-2023 inflationary cycle without lasting volume destruction.

This stock acts as a stabilizer and brand-driven defensive name as it is more of a steady compounding hold than a near-term rocket.

INTU - Intuit Inc

INTU is a quality-growth hold for the portfolio. It adds a different kind of tech exposure: recurring software, tax, and SMB-fintech rather than semis.

Revenue growth is about 15.6% YoY, profit growth is about 30.6%, ROE is about 22.8% and analysts still rate it Buy with average target around 666.8 vs current price around 422.4. The stock has been hit hard: down about 32.3% YTD, which tells us the market is worried about valuation reset and competitive pressure. Anyway, works well as a durable, high-margin software compounder.

EPD - Enterprise Products Partners L.P.

EPD operates 50,000 miles of pipeline and processing assets across the US energy infrastructure. The thesis is not a call on oil prices, it is a bet on volume throughput and the irreplaceability of the asset base. Energy demand does not disappear in any realistic medium-term scenario and building competing infrastructure at this scale is effectively impossible.

EPD gives us a different return profile: fee-based midstream cash flows, income, and lower volatility than tech stocks.

Besides that, it works a defensive energy exposure without needing oil prices to rip. It has a strong profitability with ROE around 19.2% and analyst sentiment still leans Buy.

EME - EMCOR Group

EME give us a position in industrial/infrastructure tied to construction, electrical, mechanical, data center, healthcare, and facility-services demand.

Their revenue growth is about 16.6% YoY, profit growth is about 26.4%, ROE is very strong at about 34.6% and backlog hit a record $13.25B.

It adds a high-quality cyclical compounder outside of tech, staples and energy.

CVS - CVS Health Corporation

CVS is the most speculative position in the portfolio, reflected in the smaller sizing. We do not dismiss the risks, but it adds defensive healthcare exposure that balances our higher-beta names.

The revenue growth is about 11.0% YoY, profit growth about 100.9% YoY, P/E around 11.3 and analyst consensus is Buy.

It’s a business serving 100M+ americans across pharmacy, insurance and primary care. It requires operational execution that CVS management has not yet consistently demonstrated. Small sizing, asymmetric setup.

MU - Micron Technology

Micron is one of our most aggressive positions in the portfolio, so we small sized it. It’s a high-upside cyclical growth name, not a stabilizer like KR or EPD.

They are the only US-headquartered manufacturer of DRAM and NAND memory at leading-edge scale.

The small position sizing reflects the inherent cyclicality of the business. If the HBM opportunity compounds, this becomes a larger position. For now, it earns its place as a high-conviction asymmetric bet on memory re-rating through the AI cycle.

Their revenue growth is about 48.9% YoY, profit growth is about 997.6%, ROE is about 33.3% and analysts still rate it Buy. The latest quarter was huge as they are driven by AI-related memory demand.

What comes next: Future posts will cover individual position updates when the thesis changes, new positions (with full reasoning) and macro regime shifts as we see them. No narrative management. If something goes wrong, you’ll read about it here.

Performance tracking starts from the date of this post.

The clock is running.

Disclaimer: This publication is for informational purposes only and does not constitute investment advice, a solicitation or a recommendation to buy or sell any security. The positions and opinions expressed here reflect the author's personal portfolio and views at the time of writing. Past performance is not indicative of future results. Investing involves risk, including the possible loss of principal. Always do your own research and consult a qualified financial advisor before making any investment decisions.

Share

Leave a comment

Subscribe now