AI's Pre-Halloween Massacre
First, Friday
As usual, we’ll summarize the week’s key events and how we traded them. First, Friday deserves a closer look.
Intel plunged 28.6% on October 30th, its worst day since 1974. The Nasdaq Composite fell 2.9%. A broad index can understate the damage to a concentrated holding.
Friday’s selling followed a broader retreat in hedge-fund exposure. Andreas Steno Larsen’s chart shows equity long/short funds reducing positions, or “degrossing.” Its reconstructed gross-flow series, based on Morgan Stanley Prime Brokerage data through October 27th, records its largest drawdown since the series began in 2016:
The biggest deleveraging event in more than 10 years pic.twitter.com/9FgMbMHey4
— Andreas Steno Larsen (@AndreasSteno) October 30, 2026
Protecting Yourself From Days Like Friday
Choose the largest decline you’d be comfortable taking in a stock or ETF. Enter that threshold and your share count in our optimal-hedging tool; it finds the least-expensive qualifying puts or collars. A put protects the downside; a collar helps pay for protection by selling a call, which caps upside.
Both Basic and Premium on our website include those position-hedging tools. Our iPhone app includes optimal puts; optimal collars are an optional in-app subscription. Our latest X post points to the app:
Next time, be hedged. Find the optimal puts and optimal collars to protect your stocks and ETFs with Portfolio Armor for iPhone. $INTC $AMZN
— Portfolio Armor (@PortfolioArmor) October 31, 2026
Our app starts with the position you own and the downside limit you choose. Tap the illustration to see it on the App Store:
Build Around Your Risk Limit
Across 175 completed six-month cohorts, our Top Names averaged 18.06%, versus 9.87% for SPY over the same periods.
Our Hedged Portfolio Method combines Top Names stock selection with optimal hedges for every position. Premium adds complete portfolio construction: choose an investment amount and loss limit, then get a portfolio plan designed to maximize expected return after hedging costs, including the exact share and option trades. Maintain the specified positions and hedges to keep that protection in place.
The three approaches below offer different balances of upside and risk. Their annual figures are simple illustrations from different historical records and capital bases. Trading’s 53% assumes equal-risk capital continuously redeployed at the observed return and turnover. The stock and hedged figures, 36% and 18%, double their respective six-month averages; hedged results include protection and trading costs. Tap the graphic for our guide to the approaches:
With those choices in view, here’s what drove the week’s markets and our trades.
The Cash Has To Catch Up
AI demand and the cost of serving it grew together. Amazon reported 39.3% year-over-year growth in third-quarter AWS sales. Across the company, operating cash flow reached $168.3 billion over the trailing twelve months. Net purchases of property and equipment exceeded that cash generation, leaving a $42.4 billion free-cash-flow outflow. Its quarterly results attributed the spending increase primarily to AI investment.
The market finished the week under pressure: the S&P 500 fell 1.8% Friday, the Nasdaq Composite lost 2.9%, and the Dow declined 1.1%. Growing demand still has to produce cash that can support the price investors pay.
Financing stayed central. The Federal Reserve held its target rate at 3.75%–4.00% Wednesday; two dissenters favored a quarter-point increase. For a supplier, funded projects that turn into shipments and collections matter. For a data-center operator, the timing of those collections has to support its construction and financing costs.
We examined that distinction in Monday’s neocloud financing analysis. Our trading ranged from a physical AI bottleneck to software and healthcare catalysts, with entry prices and exit plans set for each opportunity.
How We Traded It
Monday: Our Multibaggers research follows investors with documented winners of 100% or more. One of those ideas brought us back to an AI infrastructure supplier ahead of earnings. Customer commitments and improving margins supported the thesis. The bullish order filled; we closed this particular trade Wednesday. Our AI bottleneck alert explains the original setup:
Tuesday: A Market Watchers idea, drawn from investors we follow on X, proposed an earnings rebound in another AI infrastructure company. We proposed a call spread designed for an asymmetric payoff at our entry limit. The order went unfilled. Our earnings rebound alert records the proposed trade:
Wednesday: A previous software winner returned to Really Time To Buy?, our technical entry screen. We opened a new bullish trade with its own entry and exit plan. The earlier profitable exit supplied useful history; this week’s fill started a separate position. Our software winner alert lays out the renewed case:
Thursday: Biotech strength led us to three healthcare ideas. One bullish trade filled, while the other two orders were canceled unfilled and revisited Friday. Our biotech alert connects the sector backdrop to those setups:
Friday: We combined a fresh selection from our Top Names ranking with the two healthcare retries. The Top Names trade filled Friday morning; both retry orders were canceled unfilled. Our Friday alert explains the three plans:
Exits
Our completed trade on AXT illustrates how a short-term catalyst can pay before the event. We entered Monday’s call diagonal for a $2.10 net debit. The historical entry card shows the original three-contract position and its $630 maximum risk:
We exited the entire three-contract position Wednesday for a $4.10 net credit, before Thursday’s earnings. The $600 gross profit was 95% of the original $630 risk in two calendar days. The results card below records that completed trade; our separate July AXT position remained open.
Wednesday’s Taking Profits alert also distinguishes a profitable long-call sale in another position from the calendar spread that remained open.
Our weekly Exits post collects the week’s exits and commentary:
Our Year-To-Date Results
Across the 280 completed trades through October 30th, the average gross return was 20.8% on original maximum risk. There were 139 winners and 141 losers. The winners averaged 109.2%; the losers averaged −66.3%. The size of the gains relative to the losses explains the positive average with a win rate just under half.
The average holding period was 143 calendar days. Continuously deploying equal amounts of risk across trades with these average results would imply a 53.0% simple annualized return. That illustration depends on redeployment at the same average return and holding period. The dated card below summarizes the completed-trade record and links to the individual exits. Our full exits record shows the individual outcomes behind the average.
What We’re Watching Next
The October jobs report is scheduled for Friday, November 6th, at 8:30 a.m. Eastern. After the Fed’s hold and the cloud companies’ results, employment and wage data will help frame the next interest-rate decision and the financing conditions facing capital-intensive businesses.
For our next trades, we’ll keep testing funded demand, company-specific catalysts and the available option prices together. This week’s AI supplier exit shows how an opportunity can reach a worthwhile selling price before its original catalyst arrives; the unfilled orders show where our price limits kept us out.
Know someone weighing the AI buildout against its cash demands? Share this week’s review with them.
New to our alerts? Our new-entry alerts arrive before the market opens and before we enter those trades, with the investment case, option structure and exit plan. Subscribe for the actionable alerts, or subscribe free for our public posts.
Our Start Here page explains the service and includes a public sample alert: