RESEARCH & TRADE NOTES
Current thinking, documented.
Market context, portfolio decisions and lessons from execution—written before hindsight makes everything look obvious.
LATEST RESEARCH
Three recent perspectives.
NEWEST ANALYSIS
Vertiv: growth into cash, with reinvestment still to fund
Orders more than doubled and profits outpaced sales. Vertiv's cash generation and liquidity support expansion—but delivery, regional margins and the price paid still matter.
02
Meta: can the cost reset turn growth into cash?
META's advertising momentum is real, but infrastructure spending exceeds quarterly operating cash. Headcount cuts and a smaller Reality Labs loss must be judged alongside AI reinvestment and legal obligations.
Read note →
03
American Express: adjusted earnings still need footnotes
AXP's Q3 spending held up, but adjusted EPS still contains notable items. The next test is how much demand becomes sustainable pretax profit.
Read note →ARCHIVE
Earlier research.
Danaher: growth is not one number
Danaher's Q3 reported sales growth was 8.5%, versus 3% core growth. Acquisitions, a new tariff-refund exclusion and respiratory weakness explain why the bridge matters.
JPMorgan's profit grew year on year. What supported it?
JPMorgan's Q3 profit rose 6% year on year but fell 28% from Q2. A closer look at the comparison base, expenses, credit costs and the next earnings test.
Growth is not the whole story: the market decline meets TSMC's margin test
TSMC's revenue outlook challenges a simple demand-collapse story, but margins and valuation remain separate tests. What the latest market and economic evidence establishes.
Hotter inflation. No simple gold signal.
September CPI strengthens the inflation concern, while current rate pricing sharpens gold's opportunity-cost test. What the evidence changes, and what it cannot prove.
Cboe's volume is growing. What reaches earnings?
How product mix, fee capture and costs connect exchange activity to shareholder value, and the evidence needed before drawing an earnings conclusion.
Costco's September sales are strong. The profit test is still ahead
Costco's September update shows broad sales growth, even after gasoline and currency adjustments. Holiday timing, digital definitions and missing margin evidence explain why it strengthens the demand case without settling earnings or valuation.
Read the Fed minutes. Check the date: earnings and multiples for SPY and QQQ
A pre-release framework for the September Fed minutes: separate the meeting from newer PCE, jobs and services evidence, then test what changing rates mean for index earnings and valuation multiples.
Orders are growing. So are costs: the retail margin test
September's services survey still signals expansion, but rising input-price pressure complicates the earnings outlook. Costco and Target illustrate why demand, pricing and sustainable margins need separate tests.
Slower hiring, two forces: will rate relief come with resilient retail earnings?
September payrolls missed Reuters' economist poll by 61,000. For Costco and Target, softer wage pressure could help costs—but weaker hiring raises a separate question about sales and sustainable earnings.
Good news for gold. Still worth questioning
Softer-than-expected core PCE supports the case for less policy pressure. Strong real spending and flat real disposable income keep the GLD investment test open.
I own gold. I still question it: price declines, portfolio weight and GLD
Gold's drawdown is a reason to review its role, not an automatic instruction to buy. A worked GLD example separates price, portfolio weight and rebalancing discipline.
Costco beat: what lasts beyond the tariff refund?
Costco's earnings beat survives the disclosed refund adjustment. The harder investment test is whether member spending, renewal economics and operating execution justify the price paid.
Two new energy positions: how I will test XOM and APA if oil stays lower
My monthly screen produced four new positions. XOM and APA add shared oil exposure; cash generation, capital discipline and valuation must now justify the entries.
Applied Materials' $5 billion India plan: what earns the return?
Applied Materials plans a decade of investment in India. The shareholder test is how supplier capability, engineering capacity and execution become incremental cash returns.
Retail beat expectations. Will profits follow at Target?
August retail sales beat forecasts, but dollar spending is not the same as real demand or retailer profit. Target's back-to-school campaign offers a concrete test of traffic, margins and inventory.
Cboe: a trading business, not an automatic portfolio hedge
Cboe's August volumes show strength in index options alongside weaker futures activity. Why product mix, fee capture and valuation matter—and what would establish genuine diversification.
Higher fuel costs. Two different retail tests: Amazon and Ross
August's gasoline-led inflation creates different tests for Amazon's delivery economics and Ross's off-price demand. A closer look at margins, unusual earnings benefits, consumer spending and valuation.
Three stocks. One shared risk: inside the AI investment cycle
Google's €13 billion Finland plan connects computing, electrical infrastructure and mechanical systems. A research review of GOOGL, PWR and FIX: shared spending exposure, different earnings risks, and the valuation tests that matter.
Amazon's AI advantage may include the terms it can negotiate
Amazon's Qualcomm warrant highlights the bargaining power behind AI infrastructure spending. The investment test is whether better procurement economics become durable cash returns for Amazon—and profitable growth for Qualcomm.
The world economy is swerving. Optionality beats precision.
Mohamed El-Erian's transition-versus-cycle framework helps connect strategic globalization, economic statecraft and a less forgiving AI investment phase. Hormuz shows why investors need adaptability more than one perfect forecast.
Why tech can fall when oil rises
Oil shocks reach equities through two different paths: direct operating costs and the discount rate. The current question is whether Hormuz risk is a temporary premium or a durable loss of physical supply.
August recap: deploying, not trading
Eight equal-weight starters, six rule-based top-ups, no sells and cash falling from roughly 56% to 20%. August did not prove the selections right; it documented how the system handled extension, capacity and cohort timing risk.
The AI exposure hiding inside a diversified portfolio
SPY holds more than 500 companies, but its capitalization weighting concentrates exposure in a small group of AI-linked giants. My 25-position portfolio has more direct and indirect AI exposure than the sector labels suggest.
August deployment complete. Now diversification gets tested.
Eight new positions reduced cash from 58% to 20%. The important question is no longer whether the companies look different—it is whether their economic drivers remain different when Nvidia, PCE and Jackson Hole test the portfolio.
The Fed shows its hand. The consumer looks selective.
Three dissents make the July Fed minutes unusually important, while Target, Lowe's and TJX show a consumer who is still spending—but making sharper choices about where.
The consumer gets tested this week
July retail sales fell 0.6%, but five major retailer reports can show whether consumers broadly pulled back or simply shifted spending between housing, discretionary and value channels.
The week inflation cooled—and oil started climbing again
July CPI, PPI and retail sales were all soft enough to support rate hopes. Then new attacks in the Strait of Hormuz reminded markets that released data describes the past while portfolio risk lives in the present.
CPI came in cool. My calendar trick didn’t fire—and that’s the lesson.
July CPI came in broadly as expected even though oil spiked during the month. The forecast was wrong, the portfolio cost was zero, and the gap between crude oil and consumer gasoline explains why.
Inflation week—and the calendar is playing a trick on us
July CPI arrives after the oil shock it measures has partly reversed. The challenge is separating backward-looking inflation evidence from the current price path without dismissing either one.
The economy lost jobs. Stocks went up. Here’s why that isn’t crazy.
July payrolls fell while stocks rallied and Treasury yields eased. The reaction makes sense once the discount-rate benefit of weaker data is separated from the eventual earnings cost of a weaker economy.
The mood flipped. The discipline did not.
The S&P 500 returned to record territory while AMD fell after record results. The common lesson is that market mood can reverse faster than a disciplined process should.
July recap: the month the system earned its keep
July finished down 0.30% after two stop-outs, ten portfolio transactions and a violent earnings week. The result was nearly flat; the more important evidence was that the process stayed intact.
The verdict is in: AI spending needs receipts
Microsoft and Amazon were rewarded while Meta and Apple were punished. The split shows that capital spending itself is not the verdict—the market wants visible economic output.
The verdict starts landing tonight—and Teradyne is the first receipt
The Fed, Microsoft and Meta put rates and AI returns on trial tonight. Teradyne has already delivered—but a good outcome is not proof that the entry decision was right.
Last week was the warning. This week is the verdict.
Microsoft, Meta, Apple and Amazon report around a live Fed decision and June inflation data. The question is no longer whether AI is growing—it is whether the growth can justify the capital bill.
I bought on the market's worst day in a month
Ten partial positions, seven immediately underwater and 58% cash retained. This was not a prediction that the market had bottomed—it was a scheduled portfolio process operating during discomfort.
Alphabet passed the business test. The stock fell anyway.
Alphabet delivered exceptional growth across Search and Cloud, yet the stock fell sharply. The explanation sits in profit quality, capital spending and the difference between good results and expectations.
The most important earnings night of the summer—and why I’m staying flat
Alphabet and Tesla report into a market already questioning AI spending, concentrated earnings growth and rising oil. I’m not predicting the print; I’m waiting for the price it creates.
Semiconductors broke. This week tests whether the damage spreads
Semiconductors entered a bear market while Apple reached a record and small caps held up. This week's earnings will test whether that divergence represents rotation—or the beginning of broader weakness.
The inflation number everyone celebrated — and what it may be missing
June inflation cooled sharply, but much of the relief came from an energy backdrop that has already changed. Here is what CPI, PPI and bank earnings tell us—and why price structure still decides my entries.
Four Market Tests Arrive at Once — Why I’m Staying Patient
Bank earnings, inflation data, congressional testimony from Fed Chair Kevin Warsh, and renewed conflict around the Strait of Hormuz are converging in one unusually crowded market week.
What I'm watching: SpaceX index flows and the semiconductor reality check
Two live lessons in one day: why index inclusion is plumbing rather than conviction, and why great earnings are no longer enough for chip stocks priced for perfection.
My H1 2026 take: AI got expensive, gold got humbled, and the consumer is the real risk
The AI trade did not die — the invoice arrived. Why hyperscalers sold off, why memory names went vertical, why I bought Microsoft into the wreckage, and why the consumer may matter more than the AI debate in H2.
Why I'm building a quality-value sleeve alongside the breakout system
Breakout trading pays the bills, but a long-term sleeve of quality compounders bought at fair prices smooths the equity curve. Here's the framework.
The 1% rule is the whole edge
Fifteen years in, the single biggest driver of the track record isn't stock picking — it's position sizing that makes any single trade irrelevant.