2026-W3127 Jul – 31 Jul 2026

Three risk positions for a Fed week with megacap earnings in play

The screen funded nothing this cycle and the active artifact is inactive, so there is no signal to hang a directional call on. All three ideas manage exposure into the Fed decision and hyperscaler earnings rather than bet on the outcome.

Analysis as at the close of 24 Jul 2026

Wed, 29 Jul, 8:05 amNVTS and BE earnings
Thu, 30 Jul, 6:00 amFOMC rate decision and press conference
Fri, 31 Jul, 6:00 amGDP and PCE releases
Fri, 31 Jul, 8:05 amMSFT earnings, FY2027 capex guidance

Hedgethe megacap AI-capex sleeve
Hedge megacap AI exposure into the Fed and hyperscaler earnings - add protection only until the calendar clears midweek.
Thesis
The model shows zero buy candidates and the artifact producing this batch is inactive, so there is no confirmed edge to lean on this cycle. The real risk in a book carrying megacap AI exposure is concentration into a two-day gauntlet: a Wednesday Fed decision where nine of eighteen officials now project at least one hike this year against 4.2% inflation, followed a day later by Microsoft's print, where consensus centers on EPS near 4.21-4.24 and revenue near 87.7B but where the market is trading almost entirely on FY2027 capex guidance - the street wants a figure near 220B, up from roughly 190B this year, as evidence of discipline rather than a spending pace that outruns Azure's own growth. A hawkish surprise from the Fed and a capex number outside that band can move the sleeve independently of each other inside the same 48 hours. Buying protection ahead of both is a risk decision, not a directional bet.
Invalidation
Both events clear the week without a surprise, at which point the hedge should be unwound rather than rolled
Entry
add protective puts or trim beta on Monday or Tuesday, ahead of Wednesday's decision
Stop
not applicable - protective overlay, not a directional position
Size
0.25% of capital — size the hedge to the sleeve's notional exposure, not to a risk-per-trade formula
Scenarios
BULL35%−0.3R
The Fed holds with a balanced statement and Microsoft guides capex inside the expected range; the hedge cost is the only drag.
BASE40%+0.1R
One event disappoints modestly - a hawkish dot plot or capex guidance at the high end - and the hedge offsets most of the give-back.
BEAR25%+1.1R
A hawkish surprise compounds with capex guidance that reads as excess rather than discipline, and the sleeve gaps hard enough that the protection pays for itself.
Reducethe semiconductor and memory sleeve
Reduce semiconductor exposure into the earnings and rate calendar - the setup is a squeeze, not a clean re-rate. Re-add only once the dust settles.
Thesis
The signal report has no discriminating power on any name in this sleeve - every screened symbol failed the long threshold by 14 to 16 points - and the valuation cross-section shows the group split between no-value-support names (AVGO, MU, AMAT) with technical scores clustered in a neutral 46-53% band. Current reporting says DRAM pricing is up sharply on a real supply shortfall, with Micron meeting only half to two-thirds of demand, but the same names already fell hard in early July on valuation concerns before recovering, and Micron's own ATR is running near 9% of its price - wide enough that a clean directional trade this week is unusually noisy either way. Layer the Fed decision and Microsoft's capex print, which sets the read-through for AI-linked chip demand, on top of that, and the higher-beta names in the sleeve are first to move on any surprise. Trimming ahead of the calendar is a risk decision, not a bet against the memory cycle itself.
Invalidation
The Fed signals a cut or Microsoft's capex guidance is read as unambiguously bullish for AI infrastructure demand, either of which would argue for re-adding quickly
Entry
trim on Monday or Tuesday, ahead of the Wednesday decision
Stop
not applicable - exposure reduction
Size
0.25% of capital — reduce sleeve so a 10% sector gap does not break the month
Scenarios
BULL40%−0.2R
The calendar clears benignly and the memory-shortage story keeps grinding higher; the trim forgoes part of the move.
BASE35%+0.4R
A hawkish tone or a capex number read as excess hits high-beta names harder than the index, and the trim pays.
BEAR25%+1.2R
Valuation concerns resurface alongside a rate surprise and the sleeve revisits its early-July lows.
Stand asideIBM
Stand aside on IBM unless the health overlay refreshes - the margin of safety showed up because the price broke, not because the value held.
Thesis
The valuation cross-section ranks IBM as strong value support on a 37.94% margin of safety, the best reading in the current set, and the signal report's short-suitability score is a research label only, not a validated execution policy. That combination is a value-trap signature rather than an opportunity: the F-Score and F-Score v2 disagree by two points (6 vs 4), technical support is weak at 23.47%, and the valuation input date is already 25 days old against a bootstrap forecast with zero DCF fundamentals coverage across the cross-section. A high margin of safety produced by a stale, disagreeing health overlay and a model with no long-side confirmation is not a basis for a position either direction this week.
Invalidation
A refreshed valuation input date with F-Score v2 coverage and an improving technical score would put it back on the list
Entry
no entry this week at any price
Stop
not applicable - no position
Size
0.25% of capital — no capital committed
Scenarios
BULL25%−0.5R
The value case proves out and shares grind higher on the strong margin-of-safety reading; standing aside misses it.
BASE50%+0.0R
The stock ranges while the market waits on cleaner fundamental data to resolve the F-Score disagreement.
BEAR25%+0.8R
The weak technical and health readings prove more informative than the DCF screen and the stock makes new lows.

Further
analysis

Ask about any position here — the reasoning behind it, what would invalidate it, how it sits against your book. Use the contact form and we’ll come back to you.

Earlier weeks

Nothing earlier yet. The archive fills one week at a time.

This is general market research published for a general audience. It is not personalised financial advice, it does not take your circumstances into account, and it is not a recommendation to buy or sell anything. Positions described here are valid only for the stated window and are frequently wrong. Do your own work, and talk to a licensed financial adviser before acting.Disclaimer v1