Research cutoff: Friday, August 28, 2026 at roughly 12:30 BST / 07:30 ET.
U.S. price cutoff: Thursday, August 27 regular-session close.
Horizon: roughly the next 1–15 U.S. trading sessions.
Pre-open rule: extended-hours moves are context only. They do not count as confirmation for this screen.
Today's answer: Broadcom finally did something useful
Yesterday gave us the first genuinely constructive change in this screen for several days.
Broadcom (AVGO) rose 4.49% to $371.54 after Nvidia's earnings, closing above the $370 area I wanted to see reclaimed. More importantly, AVGO did not merely rise because Nvidia exploded higher. It also beat the broader semiconductor ETF SOXX, which gained 1.95%.
That matters because Nvidia itself gained 8.74%. If I use Nvidia as the only benchmark, AVGO still looks weak. If I use the broader semiconductor complex as the confirmation benchmark, AVGO finally looks like a laggard that is beginning to turn while a meaningful 10- and 20-session deficit remains.
UPS also improved. It barely moved in absolute terms Thursday, closing at $105.68, but FedEx fell 1.36%. That pushed UPS's relative performance versus FDX positive across the 1-, 3-, 5- and 10-session windows while the 20-session gap remains negative.
So the screen has two legitimate transition setups now.
I am still not calling either one a clean pre-open buy. Broadcom needs to hold its repair for a second regular session, while UPS still has not closed above the $106.20-$108 repair zone. Friday also brings Fed Chair Kevin Warsh's Jackson Hole keynote at 10:00 a.m. ET, which is a major rates catalyst for both high-multiple technology and economically sensitive industrials.
Candidate summary
| Rank | Candidate | Genuine relationship | What changed Thursday | Status | What would make it actionable |
|---|---|---|---|---|---|
| 1 | Broadcom (AVGO) | AI custom silicon/networking vs SOXX, with NVDA as the demand bellwether | AVGO rose 4.49%, reclaimed $370 and beat SOXX by ~2.54 pp while still carrying a large 10/20-session sector deficit | Closest transition | Another regular-session hold above roughly $370, preferably with AVGO flat-to-positive relative to SOXX; $375-$380 would be stronger confirmation |
| 2 | UPS | Direct parcel/logistics peer FedEx (FDX) | UPS was flat while FDX fell 1.36%, extending UPS's short-window relative turn | Improving, not broken out | Regular-session close above roughly $106.20, preferably $108+, then hold it while 3/5-session relative strength stays positive |
| — | Micron (MU) | AI memory supplier exposed to the same data-center cycle | Nvidia validated AI demand, but MU reversed from an early rally and finished weak | Rejected for now | Needs absolute stabilization and a fresh reclaim; sector-good-news underperformance is a warning, not a buy signal |
Two active watches are enough. I would rather publish two real setups than manufacture a third candidate because the format says 3-5 when available.
Relative performance through Thursday's close
The figures below are candidate price return minus benchmark price return, in percentage points. Negative means the candidate lagged its benchmark.
| Candidate vs benchmark | 1 session | 3 sessions | 5 sessions | 10 sessions | 20 sessions |
|---|---|---|---|---|---|
| UPS vs FDX | +1.39 pp | +3.67 pp | +1.37 pp | +2.47 pp | -7.02 pp |
| AVGO vs SOXX | +2.54 pp | -0.24 pp | +1.47 pp | -6.48 pp | -8.35 pp |
| AVGO vs NVDA | -4.25 pp | -5.79 pp | -3.07 pp | -12.27 pp | -21.09 pp |
That last pair of rows is the important one today.
AVGO is still far behind Nvidia, but Nvidia is the stock that just delivered the sector-moving earnings event. Requiring Broadcom to match an 8.74% Nvidia earnings jump would be too strict and would confuse leader performance with sector confirmation.
Against SOXX, AVGO now has the shape I actually want: short-window improvement, a positive catalyst-day relative move, and a still-unresolved medium-window deficit.
Seven-session walk-forward: one benchmark was not enough
I re-ran the screen across the prior seven completed U.S. sessions using only the information that would have been available at each cutoff.
I am keeping the raw hit rate private. The useful output is the rule change.
The recent sample showed three different failure modes:
- RIOT produced sharp one-day relative bounces that failed immediately. The second-session rule prevented those from becoming completed signals.
- CDNS eventually caught up so quickly that the old 20-session deficit became stale. The freshness filter prevented us from chasing a move that had already happened.
- AVGO exposed a benchmark problem. On a major Nvidia earnings day, comparing AVGO only with Nvidia makes a healthy sector response look like failure because Nvidia itself is the event leader.
New refinement: leader + basket confirmation
When one company is the direct catalyst for an entire industry, I now use two benchmarks:
- the catalyst leader, to measure how much catch-up room remains; and
- a diversified sector basket or ETF, to decide whether the laggard is actually participating.
A candidate does not need to beat the catalyst leader on the event day. It does need to show positive absolute price action and preferably outperform the diversified basket.
This is a material change, not a wording tweak. It stops the screen from rejecting a valid transition simply because the industry leader had an unusually large earnings move.
There is one more wrinkle today: Marvell is down sharply in Friday premarket trading after earnings, even though it raised longer-term revenue forecasts. That can drag semiconductor baskets such as SOXX lower. Under the existing peer-shock rule, any apparent AVGO outperformance versus SOXX in Friday premarket is therefore ignored. Regular-session confirmation has to survive that contamination.
1. Broadcom: yesterday passed the first real test
Why this relationship is genuine
Broadcom is not simply "another Nvidia."
Nvidia sells merchant AI accelerators. Broadcom's AI exposure is concentrated in custom accelerators, networking and hyperscaler infrastructure, alongside a large infrastructure-software business.
The shared economic driver is hyperscaler AI capital expenditure.
Nvidia's latest quarter reinforced that driver. Reuters reported that Nvidia's forecast revived confidence in the AI trade, while chip-related stocks rallied across the sector. The iShares Semiconductor ETF rose 1.95% Thursday.
Broadcom did better than that.
The fundamentals remain strong
Broadcom's fiscal Q2 revenue was $22.19 billion, up 48% year over year. AI semiconductor revenue reached $10.8 billion, up 143%, and management guided fiscal Q3 AI semiconductor revenue to roughly $16 billion.
That is not a broken AI business.
The risk is expectations. Broadcom has already shown that spectacular AI growth can still disappoint a market priced for even more. Customer concentration, custom-chip competition and financing/debt sensitivity remain real reasons for the stock to trade differently from Nvidia.
What Thursday changed
AVGO closed at $371.54, up 4.49%.
SOXX gained 1.95%.
That gives AVGO a +2.54 percentage-point one-session relative win versus the diversified semiconductor basket. It also leaves AVGO roughly 6.5 points behind SOXX over 10 sessions and 8.3 points behind over 20.
That is a much cleaner catch-up shape than yesterday's premarket setup.
Against Nvidia, AVGO still trails badly. I view that as remaining upside room, not evidence by itself that AVGO failed Thursday's sector test.
| Item | AVGO plan |
|---|---|
| Observation area | Roughly $365-$372 |
| First trigger | Already achieved: Thursday close above $370 |
| Confirmation | Another regular-session close holding roughly $370, preferably $375-$380, without materially underperforming SOXX |
| Invalidation | A quick close back below roughly $360 would weaken the catalyst-turn thesis; below $350 would be a stronger rejection |
| Likely catalysts | Continued AI estimate revisions, custom-silicon/networking demand, OpenAI/Google/Anthropic deployments, Sept. 2 Broadcom earnings |
| Main risks | Customer concentration, hyperscaler diversification, valuation, financing/debt sensitivity, semiconductor tariffs, earnings event risk |
Sell-side sentiment remains strongly constructive. Morgan Stanley, KeyBanc and BNP Paribas Exane have all published bullish targets after Broadcom's latest quarter, although UBS trimmed its target slightly while keeping a Buy rating because of supply constraints.
Retail sentiment is more mixed, which I prefer. A recent Broadcom-focused Reddit thread discussed the stock's unusually large swings versus Nvidia and repeatedly raised debt, institutional ownership and customer concentration as reasons not to assume every dip is mispricing.
Verdict: AVGO is now the closest thing on the screen to a real catch-up transition. I still want the second regular session. One catalyst day is evidence; two sessions make it much harder to dismiss as sympathy buying.
2. UPS: the relative turn is now broad, but price still has not escaped
Why UPS versus FedEx is one of the cleanest pairs
UPS and FedEx compete directly in parcel delivery, business shipping, international logistics and time-sensitive freight. They share exposure to economic activity, fuel, labor, trade volumes, e-commerce and network utilization.
That makes relative divergence meaningful when company-specific explanations do not dominate.
The business case remains intact
UPS reported $22.8 billion of Q2 revenue and $1.76 adjusted EPS, and raised its full-year outlook to roughly $91.2 billion of revenue, $8.65 billion of adjusted operating profit and $7.22 adjusted EPS.
On August 24 the company also announced more than $2 billion of ongoing investment across its International, Healthcare and Supply Chain Solutions businesses.
The bear case has not disappeared. The Amazon volume reset, restructuring costs, weak parcel volumes and international margin pressure all remain legitimate concerns.
That is why the stock needs to prove itself rather than being bought simply because FedEx has run farther.
What Thursday changed
UPS closed at $105.68, up only 0.03%.
FedEx fell 1.36%.
That was enough to push UPS ahead of FDX over the 1-, 3-, 5- and 10-session windows, while the 20-session gap remains around -7.0 percentage points on raw price returns.
The shape is excellent for a catch-up screen.
The absolute chart still refuses to confirm it.
UPS traded as high as $106.25 Thursday but closed back at $105.68, almost exactly where it started. That is why I am not moving the trigger downward just to call the setup complete.
| Item | UPS plan |
|---|---|
| Observation area | Roughly $104-$106 |
| First useful trigger | Regular-session close above roughly $106.20 |
| Better confirmation | Push through $108, then hold the area while 3/5-session relative performance versus FDX stays positive |
| Invalidation | Fresh close below roughly $101-$102 materially weakens the transition |
| Likely catalysts | Better volume quality after the Amazon reset, healthcare/international expansion, network-cost leverage, supportive estimate revisions |
| Main risks | Weak parcel demand, restructuring costs, labor/fuel pressure, tariffs, Amazon and FedEx competition |
Wall Street is constructive but not unanimous. StockAnalysis shows a Buy consensus, with recent August views ranging from Bernstein Buy to BMO and Wells Fargo Hold. That spread fits the setup: there is upside if the turnaround works, but there is still enough skepticism to keep the stock from already pricing in success.
Community discussion is similarly split. Recent UPS employee and investor threads point to both real network/volume weakness and signs that the Amazon reset is more nuanced than the headline "50% reduction" suggests.
Verdict: UPS has the right relative pattern. I still want the price breakout. The whole point of this screen is to buy a turn early, not to buy a statistical divergence while the stock itself refuses to move.
Why Micron is not today's third pick
Micron is exactly the kind of stock that could tempt this strategy into a bad shortcut.
Nvidia's results reinforced extraordinary AI-memory demand and warned about tight memory conditions. Micron's own recent numbers are extremely strong.
Yet MU reversed from an early Nvidia-driven rally on Thursday and traded lower while the broader semiconductor complex finished strongly positive.
Recent coverage attributed the weakness partly to profit-taking after a huge year-to-date run and to the uncomfortable fact that memory scarcity is both good for pricing and potentially bad for system shipment volumes. Valuation expectations are also extremely high.
That is not unexplained relative weakness.
Verdict: rejected for now. I want absolute stabilization first. A stock that cannot hold a rally on one of the best possible industry-demand days should not be promoted simply because its peer went up more.
Friday macro backdrop: the 10 a.m. ET speech matters more than premarket
U.S. futures are mixed Friday morning after Thursday's technology-heavy rally. The Nasdaq 100 is softer while the Dow is slightly firmer.
The main macro event is Fed Chair Kevin Warsh's Jackson Hole keynote at 10:00 a.m. ET. The Federal Reserve's official calendar confirms the timing.
That speech matters to both names:
- AVGO is sensitive to long-duration growth-stock valuation and Treasury yields.
- UPS is sensitive to the economic-growth outlook, financing conditions and business activity.
There is also a fresh semiconductor-specific distortion: Marvell is down roughly 8% in premarket trading after investors focused on the timing of revenue from its Google AI deal despite strong results and raised long-term forecasts.
That means Friday's premarket SOXX move is not a clean benchmark for AVGO. I will not count it.
Bottom line
This is the first day in a while where the screen has two genuinely improving setups rather than two statistical laggards.
Broadcom is closest. It reclaimed $370, beat the broader semiconductor basket on the Nvidia catalyst day, and still has meaningful 10/20-session catch-up room. One more regular session holding the repair would materially strengthen the case.
UPS is next. Its relative pattern versus FedEx is now positive across every window except 20 sessions, but the stock still has not closed above the repair zone.
So I am not forcing a buy before the bell.
The levels I care about are simple:
- AVGO: hold $370, preferably push into $375-$380 with healthy sector-relative action.
- UPS: close above $106.20, preferably $108+, without giving back the short-window relative turn.
If those conditions survive Friday's Jackson Hole volatility, Monday's screen should finally have something stronger than a watchlist.
Sources
Market, semiconductors and macro
- Reuters, Aug. 27 — Nvidia forecast revives AI trade and lifts chip stocks: https://www.reuters.com/business/nvidia-rises-after-signaling-longer-ai-spending-runway-2026-08-27/
- Reuters, Aug. 27 — Nasdaq and S&P lifted by Nvidia; investors await Warsh: https://www.reuters.com/business/nasdaq-futures-take-lead-after-nvidia-forecast-refuels-ai-trade-2026-08-27/
- Reuters, Aug. 28 — Warsh's Jackson Hole speech and policy expectations: https://www.reuters.com/business/will-warshs-jackson-hole-speech-be-course-correction-or-detour-2026-08-28/
- Federal Reserve calendar — Warsh keynote at 10:00 a.m. ET: https://www.federalreserve.gov/newsevents/2026-august.htm
- Reuters, Aug. 28 — Marvell falls as investors question timing of Google AI-deal revenue: https://www.reuters.com/business/marvell-shares-slide-concerns-over-timing-google-ai-deal-revenue-eclipse-strong-2026-08-28/
Price data
- Broadcom price history: https://stockanalysis.com/stocks/avgo/history/
- Nvidia price history: https://stockanalysis.com/stocks/nvda/history/
- SOXX price history: https://stockanalysis.com/etf/soxx/history/
- UPS price history: https://stockanalysis.com/stocks/ups/history/
- FedEx price history: https://stockanalysis.com/stocks/fdx/history/
Broadcom
- Broadcom fiscal Q2 2026 results: https://investors.broadcom.com/news-releases/news-release-details/broadcom-inc-announces-second-quarter-fiscal-year-2026-financial
- Investing.com — Morgan Stanley raises AVGO target on AI revenue outlook: https://www.investing.com/news/analyst-ratings/morgan-stanley-raises-broadcom-stock-price-target-on-ai-revenue-outlook-93CH-4725935
- Investing.com — KeyBanc raises AVGO target on AI outlook: https://www.investing.com/news/analyst-ratings/keybanc-raises-broadcom-stock-price-target-to-575-on-ai-outlook-93CH-4726588
- Reddit, Aug. 26 — Broadcom volatility/debt discussion: https://www.reddit.com/r/BroadcomStock/comments/1vyz74c/why_does_avgos_price_appear_to_swing_more_than/
UPS
- UPS Q2 2026 results and raised guidance: https://about.ups.com/us/en/newsroom/press-releases/financials/ups-releases-2q-2026-earnings.html
- UPS, Aug. 24 — more than $2B International/Healthcare/Supply Chain investment: https://about.ups.com/us/en/newsroom/press-releases/customer-first/ups-invests-more-than--2-billion-to-give-customers-even-faster-s.html
- StockAnalysis — UPS analyst consensus and recent ratings: https://stockanalysis.com/stocks/ups/forecast/
- Reddit — UPS Q2 earnings discussion: https://www.reddit.com/r/EarningsCalls/comments/1va6zrd/ups_ups_the_good_the_bad_and_the_ugly_from_upss/
Micron rejection context
- Yahoo Finance / GuruFocus, Aug. 27 — Micron's Nvidia rally fades amid memory-scarcity concerns: https://finance.yahoo.com/markets/stocks/articles/microns-nvidia-rally-vanishes-memory-193128746.html
- Benzinga, Aug. 27 — profit-taking and valuation concerns around MU: https://www.benzinga.com/trading-ideas/movers/26/08/61474888/why-is-micron-stock-falling-on-thursday
This is market research, not individualized financial advice. Relative gaps can stay open far longer than expected; predefined invalidation and position sizing matter more than the elegance of the catch-up story.