Daily Tech Report

Research cutoff: Sunday, September 6, 2026, roughly 08:30 BST.
U.S. price cutoff: Friday, September 4 regular-session close, plus confirmed weekend news through early Sunday.
Horizon: roughly the next 1–7 U.S. trading sessions.
Weekend note: U.S. cash markets are closed today and are also closed Monday, September 7 for Labor Day. The next regular U.S. session is Tuesday, September 8. U.S. index futures have not yet given us a useful Sunday-morning signal, so this is a Tuesday/next-week playbook, not a fake live pre-market update.

The quick read

Friday gave tech investors a strange but useful signal.

The broad market fell after the August U.S. jobs report came in much hotter than expected: 162,000 payrolls versus 56,000 expected, with unemployment holding at 4.1%. The S&P 500 fell 0.38%, the Nasdaq lost 0.29%, and the Dow dropped 0.51%. Market-implied odds of a 25 bp September Fed hike rose to roughly 58%.

But underneath that weak index close, semiconductors surged 3.4%.

That is the part I care about most going into Tuesday.

AMD jumped 4.69% to $477.57, the strongest stock in this nine-name group.
NVDA gained 0.84% to $230.36, holding green despite the rates shock.
META added 1.00% to $616.77, extending its breakout above $600.
AVGO edged up 0.21% to $357.90, stabilizing after its post-earnings selloff.
MSFT fell 2.04%, AAPL fell 2.51%, and GOOGL fell 1.17% as software/platform megacaps were more exposed to the yield repricing.
TSLA dropped 5.92% to $354.08 after the Cybercab excitement collided with a fresh NHTSA probe.

Then the weekend added two opposing forces.

First, the macro risk got worse: the U.S. military said it struck three Iranian crude-oil carriers on Saturday after Iran fired ballistic missiles at U.S. Navy ships. There is no valid Sunday-morning oil or Nasdaq-futures price to quote yet, but the direction of the risk is clear: oil could reopen higher, feeding inflation expectations and Treasury yields before Tuesday's stock-market open.

Second, the AI-demand data got stronger. Foxconn said third-quarter performance should beat market expectations on AI strength, with August revenue up 51.98% year over year to a record-for-August T$921.8 billion. South Korea also reported extraordinary AI-driven semiconductor exports: January-August chip exports rose 169.6% to about $281 billion, representing roughly 41% of total exports.

So the setup is not simply bullish or bearish.

My interpretation: AI infrastructure demand is still exceptionally strong, but the discount-rate problem is getting harder. I want to favor stocks proving relative strength, avoid chasing Friday's vertical winners, and make the laggards earn a catch-up label by reclaiming nearby levels.

Market pulse going into the holiday-shortened week

Market / macro	Latest confirmed read	What it means for tech
S&P 500, Friday	-0.38%	Hot payrolls pressured the broad market
Nasdaq Composite, Friday	-0.29%	Mild index decline hid a major sector split
Dow Jones, Friday	-0.51%	Broad risk appetite weakened
Semiconductor group	+3.4% Friday	Clear relative-strength pocket
August payrolls	+162k vs +56k expected	Labor market much stronger than priced
Unemployment	4.1%	No deterioration signal from the headline rate
September Fed hike odds	~58% Friday	Up sharply after payrolls
U.S. 2Y Treasury	~4.37% Friday	Direct read on Fed expectations
U.S. 10Y Treasury	~4.77–4.78% Friday	Major valuation pressure for long-duration tech
Brent crude, Friday	~$92.68/bbl	Already elevated before fresh weekend escalation
Weekend geopolitical update	U.S. struck 3 Iranian crude carriers	Oil/inflation gap risk when markets reopen
Monday, Sept. 7	U.S. cash market closed	Next regular session is Tuesday
PPI	Thu, Sept. 10, 08:30 ET	First major inflation test of the week
CPI	Fri, Sept. 11, 08:30 ET	Likely the biggest macro catalyst before the Fed

The key macro tension is straightforward. Friday's jobs number already gave the Fed more room to tighten. If the weekend conflict pushes oil materially higher, the market can end up with strong labor + higher energy + already-elevated yields at the same time.

That is an uncomfortable combination for expensive technology stocks even if the earnings story is excellent.

The tech board

Stock	Friday close	Friday move	1–7 day read
NVDA	$230.36	+0.84%	Strong relative tape + Hugging Face catalyst + fresh AI-demand checks
MSFT	$499.70	-2.04%	High-quality laggard; $500 area is now the key decision point
AAPL	$319.97	-2.51%	Weak Friday, but Sept. 9 event creates a near-term catalyst
GOOGL	$338.46	-1.17%	Legal de-risking is positive; price still needs repair
META	$616.77	+1.00%	Cleanest megacap relative-strength trend
AMZN	$258.51	-0.15%	Resilient versus software peers; legal overhang remains
TSLA	$354.08	-5.92%	Cybercab momentum flipped into regulatory-risk trade
AMD	$477.57	+4.69%	Friday's strongest stock; now directly at resistance/chase-risk area
AVGO	$357.90	+0.21%	Stabilizing after earnings; repair still incomplete

The levels below are observation/decision zones derived from recent regular-session highs, lows and closes. They are not analyst targets, guarantees or automatic entries.

The weekend macro change: oil is now Tuesday's first chart

Friday already ended with Brent around $92.68 and the U.S. 10-year Treasury around 4.77%–4.78%.

Then Saturday brought a new escalation. U.S. Central Command said U.S. forces struck three Iranian crude-oil carriers after ballistic missiles were fired at two U.S. Navy ships. No U.S. personnel were reported injured.

I am deliberately not putting a Sunday oil price in this report because there is no meaningful live Sunday-morning quote yet.

The trade implication is conditional:

If crude gaps higher when futures reopen, inflation expectations can rise again.
If the 10-year and especially the 2-year move higher with oil, Friday's software weakness can continue.
If oil absorbs the news without a major move, then Friday's semiconductor relative strength becomes much easier to trust.

For Tuesday, I would check oil and Treasury yields before looking at NVDA or AMD. The best chip chart in the world can still get hit if the macro discount rate jumps another leg.

The fresh AI-demand check is genuinely strong

The weekend gave us two independent pieces of evidence that the AI-hardware cycle is still running hot.

Foxconn: record August revenue and an AI-driven Q3 beat setup

Foxconn said it expects third-quarter performance to exceed market expectations because of strong AI-product demand and seasonal strength in information/communications products.

August revenue reached T$921.8 billion ($29.15 billion), up 51.98% year over year, the highest August total on record and the second consecutive month above T$900 billion.

Foxconn matters because it is Nvidia's largest server manufacturer and a major Apple supplier.

This does not mean every dollar of Foxconn revenue becomes Nvidia revenue. It does mean the physical manufacturing layer is confirming the same demand signal we have been hearing from Nvidia, Dell and Broadcom.

South Korea: AI demand is showing up in national export data

South Korea's year-to-date exports reached $709.4 billion, already surpassing the previous full-year record.

The extraordinary part is semiconductors: January-August chip exports rose about 169.6% to $281 billion, accounting for roughly 41% of total exports.

Again, this is industry data rather than an AMD/Nvidia earnings report. But it is useful because it is difficult to argue that AI infrastructure demand is merely a management-guidance story when the same boom is showing up in server-manufacturing revenue and national trade flows.

My interpretation: the AI-cycle fundamental risk is currently less about demand disappearing and more about valuation, supply, custom-silicon competition, execution timing and interest rates.

Nvidia: the strongest combination of tape, fresh demand and strategic optionality

Nvidia closed Friday at $230.36, up 0.84%, after trading between roughly $229.63 and $234.76.

That was a very good relative performance on a day when the Nasdaq fell and the jobs report pushed rate-hike odds higher.

The company also has one of the week's biggest strategic AI developments behind it: Nvidia agreed to acquire Hugging Face for $12.93 billion. Hugging Face is one of the most important distribution and developer platforms for open AI models.

The strategic logic is bigger than another revenue stream. Nvidia already dominates the compute layer. Hugging Face moves it closer to the point where developers discover, download, fine-tune and deploy models.

Nvidia says Hugging Face will remain open and interoperable, which is essential because the platform loses much of its value if developers see it as simply an Nvidia lock-in mechanism.

Still, the acquisition gives Nvidia another defensive layer as Meta, Microsoft, Google, Amazon and OpenAI all invest in custom accelerators.

There is also a PC expansion angle: Lenovo and Acer are set to launch the first Windows PCs using Nvidia's RTX Spark AI chip in October.

NVDA decision zones

$229–230: immediate support/acceptance area from Friday's low and close.
$234.75–235: Friday's high and immediate breakout test.
$224–225: deeper support around Thursday's low / prior breakout area.

My read: probably the best combination of company catalyst and price confirmation. I would rather see a controlled $229–230 retest hold or a clean $235 breakout than chase a Tuesday gap after weekend headlines.

AMD: Friday's strongest stock — which also makes it the easiest one to chase badly

AMD jumped 4.69% to $477.57 Friday, after opening around $462 and trading as high as $478.83.

The move matters because AMD had been the obvious semiconductor laggard earlier in the week. Friday changed that.

Part of the move was sector-wide: semiconductors rose 3.4%. But AMD still outperformed a very strong chip tape.

The fundamental backdrop remains supportive. AMD recently won a role in Europe's LUMI-AI supercomputer, and the weekend's South Korean semiconductor-export data reinforces the broader hardware-demand cycle.

The problem is entry quality.

AMD is now sitting right at a cluster of recent highs after a nearly 5% one-day move.

AMD decision zones

$458–462: first deeper support/retest area around Friday's low/open and the prior reclaim zone.
$478–483: immediate resistance. Friday topped at $478.83 and the stock recently traded around $483.
$491: larger swing-high area if $483 breaks cleanly.

My read: strongest Friday momentum, worst FOMO risk. I prefer a clean $483 break with sector confirmation or a controlled pullback. I do not like buying simply because Friday was green.

Meta: $600 is finally behaving more like support than resistance

Meta closed at $616.77, up 1.00% Friday after Thursday's 3% breakout above $600.

Friday's range was roughly $605.27 to $617.46, which is useful: buyers stepped in above the prior breakout zone even while the broader market fell.

There was no single Friday Meta headline that explains the entire move, and I would rather say that than invent one.

What we do know is that Meta remains one of the largest AI-infrastructure spenders in the world. Broadcom has cited Meta as a major custom-chip/networking customer, and Reuters reported earlier this summer that Meta planned to put its own AI chip into production in September while working to roughly double computing capacity by 2027.

That gives Meta an unusual position: it is both a massive buyer of AI infrastructure and increasingly a designer of its own stack.

META decision zones

$610–611: first support/acceptance area around Thursday's close.
$604–605: deeper pullback support around Thursday/Friday lows.
$619–620: immediate breakout test around Thursday's high.

My read: the cleanest megacap trend on the board. The only issue is entry quality after a sharp multi-day run. A successful $610 retest is more attractive to me than chasing a gap through $620.

Broadcom: spectacular AI fundamentals, unfinished technical repair

Broadcom closed $357.90, up 0.21% Friday after the stock sold off on its earnings reaction.

The long-term AI numbers remain extraordinary:

Fiscal 2027 AI-chip revenue forecast around $115 billion.
Fiscal 2028 AI-chip revenue expected around $230 billion.
Q3 AI-chip revenue of roughly $16.7 billion.
AI-chip bookings above $30 billion in the quarter.

The reason the stock did not explode higher was the near-term bar. Broadcom guided Q4 revenue to about $34.8 billion, a little below the Reuters/LSEG consensus cited around the report.

Friday's range was $353.70 to $360.16.

So AVGO did something useful — it stopped falling — but it still badly lagged AMD and the broader semiconductor group.

AVGO decision zones

$353–355: immediate support/stabilization area.
$359.5–360: first reclaim level around Friday's high.
$364–367: more meaningful repair zone from the pre-earnings range.
$342–345: earnings-reaction low zone if stabilization fails.

My read: one of the better catch-up candidates, but not yet a confirmed leader. I want $360 first and preferably $364–367 before saying the earnings gap has been absorbed.

Tesla: Cybercab is now a regulatory trade, not just an autonomy-hype trade

Tesla fell 5.92% to $354.08 Friday after gaining 5.42% Thursday around the Cybercab event.

Friday's range was roughly $351.32 to $364.69.

The reversal was not just generic profit-taking. The U.S. National Highway Traffic Safety Administration opened a probe into whether Tesla properly self-certified up to 1,000 Cybercabs under federal safety standards.

The issue matters because the purpose-built Cybercab lacks conventional controls such as a steering wheel and pedals. Reuters reported Tesla had not followed the exemption route used by some autonomous-vehicle developers.

Tesla has started limited rides in Austin and has real vehicles on the road, so the story is not theoretical. But the regulatory path to scaled commercial deployment is now part of the valuation debate.

The comparison with Amazon's Zoox is worth noting: Zoox previously received a federal path for limited paid robotaxi deployment without normal human controls. That does not mean Tesla cannot get there. It means the regulatory process matters.

TSLA decision zones

$351–354: immediate support around Friday's low/close.
$362–365: first repair zone around Friday's open/high area.
$376: larger reclaim level — Thursday's close before the probe-driven selloff.

My read: highest company-specific headline risk in the group. I would not buy just because it dropped 6%. If $351–354 holds and Tesla reclaims $362–365, the trade can stabilize. If not, Friday's reversal deserves respect.

Microsoft: best quality catch-up candidate, but $500 now matters

Microsoft closed $499.70, down 2.04%, after trading between roughly $499.36 and $511.00.

That was poor relative performance compared with the chip group, but it also happened during a day when higher rates hit software and duration-sensitive growth harder.

The underlying cloud catalyst from earlier in the week is still meaningful. Microsoft disclosed Azure revenue more directly for the first time: $29.4 billion in the latest quarter and $101.9 billion for the fiscal year.

That gives investors a clearer view of the actual scale of Azure rather than only a percentage growth rate.

There is fresh weekend legal noise too. The Seattle Times and Newsday sued OpenAI and Microsoft on Friday, alleging copyright infringement in AI training. Those are allegations, not findings of liability, and I would treat the case as a sentiment/legal overhang rather than a one-week earnings reset.

MSFT decision zones

$499–500: immediate support/psychological decision area.
$510–511: first meaningful repair zone around Friday's open/high.
$515–518: stronger breakout/reclaim area from recent highs.
$493–496: deeper support if $500 fails.

My read: high-quality catch-up candidate, not current relative strength. If yields settle and MSFT reclaims $510–511 quickly, the trade improves a lot. If it lives below $500 while chips keep leading, I would not force it.

Apple: weak Friday, but Wednesday is now close enough to matter

Apple fell 2.51% to $319.97 Friday after trading between roughly $317.86 and $328.93.

That was one of the weakest megacap performances outside Tesla.

But Apple has the clearest scheduled company catalyst in this group: its official "Surprise and shine" Apple Event is Wednesday, September 9 at 6 p.m. BST / 10 a.m. PT.

Apple itself confirms the event timing. Reuters expects the next iPhone lineup to be the focus and says investors are also watching for a possible foldable device. The event is confirmed; the exact product mix remains expectation rather than fact until Apple announces it.

Foxconn's weekend update is a secondary positive read-through because Foxconn is a major Apple supplier, although its current growth is being driven heavily by AI-server demand rather than only consumer electronics.

AAPL decision zones

$318–320: immediate support around Friday's low/close.
$329–331: first repair/breakout area from Friday/Thursday highs.
$313–315: deeper support from recent repeated lows.

My read: this is an event setup, not a pure momentum trade. If $318–320 holds into Wednesday, the risk becomes easier to define. A reclaim of $329–331 would tell us buyers are positioning positively into the launch.

Google: good legal news, but the stock still has to prove it

Alphabet Class A closed $338.46, down 1.17% Friday.

The legal backdrop improved earlier in the week when a U.S. judge rejected the Justice Department's request to force Google to sell its AdX advertising exchange.

That does not mean Google was cleared of antitrust violations. The positive is specifically that the remedy avoided a forced structural breakup of AdX.

European scrutiny is still active, including questions around publishers' ability to opt out of AI-generated search features.

So GOOGL currently has a classic setup where the narrative improved more than the chart.

GOOGL decision zones

$337–338: immediate support around Friday's low/close area.
$342–344: first repair zone around recent closes/highs.
$349–350: larger breakout/repair area from late August.

My read: catch-up candidate, not leader. If yields calm and GOOGL still cannot reclaim $342–344, I would downgrade the idea. A quick reclaim would make the legal de-risking much more relevant to the trade.

Amazon: quietly resilient, but still carrying legal baggage

Amazon closed $258.51, down only 0.15%, after trading between roughly $255.29 and $261.12.

That is a respectable result given Friday's software/platform weakness.

The company still has the FTC and 22-state lawsuit over advertising-auction practices hanging over it. Amazon denies the allegations. For a 1–7 day trade, the important question is whether the stock can absorb the legal overhang and recover its pre-lawsuit range.

The AI/cloud backdrop remains favorable. AWS is one of the largest direct beneficiaries of the compute buildout, and the weekend Foxconn/South Korea data give another external confirmation that the physical infrastructure cycle is strong.

AMZN decision zones

$255–256: immediate support around Friday's low.
$261–262: first breakout/reclaim area.
$264–266: larger repair zone from the pre-lawsuit range.

My read: resilient but lower urgency. Above $261–262 the setup improves; above $264–266 I would be much more comfortable saying the legal-news damage is being absorbed.

Relative strength: Friday changed the hierarchy

Among the nine names, Friday's ranking was approximately:

AMD +4.69%
META +1.00%
NVDA +0.84%
AVGO +0.21%
AMZN -0.15%
GOOGL -1.17%
MSFT -2.04%
AAPL -2.51%
TSLA -5.92%

I would split them into four buckets.

Confirmed relative strength: AMD / NVDA / META

All three beat the Nasdaq by a wide margin Friday.

AMD has the strongest momentum and the highest chase risk.
NVDA has the best combination of fresh company catalyst and external AI-demand confirmation.
META has the cleanest megacap trend.

Stabilizing: AVGO / AMZN

Broadcom stopped the earnings-reaction decline but still lagged the chip rally. Amazon held up better than the software megacaps but still has a company-specific legal overhang.

Catch-up candidates: MSFT / GOOGL / AAPL

A laggard is not automatically a bargain. I want actual price repair:

MSFT: $510–511.
GOOGL: $342–344.
AAPL: $329–331, with the September 9 event as a catalyst.

Event/regulatory risk: TSLA

Tesla currently has its own market. The Cybercab certification probe can overwhelm whatever the Nasdaq does.

The next week's catalyst calendar

Because Monday is a market holiday, the week is compressed.

Tuesday, September 8 — first market reaction to the weekend

This is the first regular session after both the hot jobs report and the weekend U.S.-Iran escalation.

The first three things I would check are:

Brent/WTI crude.
The U.S. 2-year and 10-year Treasury yields.
Semiconductors versus software/platform megacaps.

If yields stay high but AMD/NVDA keep outperforming, Friday's chip strength has more credibility.

If yields ease and MSFT/GOOGL/AAPL reclaim nearby levels, the more attractive trade may become catch-up rather than chasing semiconductors.

Wednesday, September 9 — Apple event

Apple's official event begins at 10 a.m. PT / 1 p.m. ET / 6 p.m. BST.

For AAPL, the reaction matters more than predicting exactly which products Apple unveils. The important trading question is whether investors buy the event, sell the news, or use it to reprice expectations under the new leadership team.

Thursday, September 10 — PPI + Oracle earnings

The BLS releases August PPI at 8:30 a.m. ET.

This is the first major inflation test after Friday's hot payroll report and after the weekend oil escalation.

Oracle also reports earnings Thursday. ORCL is not one of the nine core stocks here, but its results matter for the broader AI/cloud trade because of its expanding role in AI data-center capacity and large model-company infrastructure contracts.

A strong Oracle cloud/AI read can provide a sympathy signal for NVDA, AVGO, MSFT, AMZN and the infrastructure ecosystem.

Friday, September 11 — CPI at 8:30 a.m. ET

This is probably the week's most important macro event.

The BLS confirms the August CPI release for 8:30 a.m. ET. Reuters' current economist polling has headline CPI around +0.4% month over month and core around +0.2%.

The Fed meets September 15–16, so CPI lands immediately before the decision window.

The market has already moved from treating a September hike as a tail risk to roughly a coin-flip-plus probability. A hot CPI, especially with higher oil, can push that probability much higher. A cooler print can relieve pressure quickly.

What I'd watch into Tuesday — ranked

This is a ranking of information-rich setups, not nine instructions to buy.

1. NVDA — best combination of catalyst, tape and external demand confirmation

Friday held green through a hawkish jobs shock. The Hugging Face acquisition expands the platform story, and Foxconn/South Korea data reinforce the underlying AI cycle. I care about $229–230 support and $234.75–235 breakout.

2. AMD — strongest Friday momentum, highest chase risk

A +4.69% day in a falling Nasdaq is serious relative strength. But $478–483 is now a decision area, not an easy entry. I prefer confirmation or a controlled pullback.

3. AAPL — the cleanest scheduled company catalyst

Wednesday's event is close enough that positioning can dominate the next two sessions. $318–320 needs to hold; $329–331 is the first meaningful repair.

4. TSLA — highest company-specific information value

The story changed from Cybercab excitement to certification/regulatory risk within 24 hours. $351–354 is the first hold zone; $362–365 is the first repair. I would not guess the bottom.

5. AVGO — best semiconductor catch-up test

Broadcom's long-term AI numbers are exceptional, but the stock has not confirmed them. $360 first, then $364–367 is what I want back.

6. META — cleanest megacap trend

The $600 breakout survived a difficult macro day. $610–611 is the first pullback zone; $619–620 is the next momentum test.

7. MSFT — best quality laggard/catch-up idea

I want $500 to hold and then $510–511 reclaimed. If yields cool and Microsoft cannot recover, that is useful negative information.

8. GOOGL — legal de-risking needs price confirmation

The AdX breakup risk improved, but the stock still lagged Friday. $337–338 support and $342–344 repair keep the setup simple.

9. AMZN — resilient, lower urgency

Amazon is behaving better than MSFT/AAPL on the downside, but the FTC case still creates an overhang. $261–262 and then $264–266 are the levels that would make the repair more convincing.

The 1–7 session playbook

Scenario 1: oil gaps higher and yields rise again

This is the main new weekend risk.

In that environment, I would not assume Friday's semiconductor strength automatically continues. I would watch which leaders can hold support despite a worse discount-rate backdrop.

The strongest information would be:

NVDA holds $229–230.
META holds $610.
AMD holds the $458–462 reclaim zone on a deeper pullback.

If those levels fail immediately while yields surge, the macro is winning.

Scenario 2: oil absorbs the weekend escalation and yields stabilize

This is the friendlier setup for equities.

Then I would give more weight to Friday's chip leadership and the fresh AI-demand data.

NVDA above $235 becomes a cleaner continuation trade.
AMD above $483 confirms the Friday breakout rather than merely extending it emotionally.
AVGO above $360, then $364–367, becomes the catch-up candidate.

Scenario 3: yields cool and software/platform names catch up

This is where MSFT, GOOGL and AAPL become much more interesting than buying the top of AMD's Friday candle.

I want price confirmation:

MSFT above $510–511.
GOOGL above $342–344.
AAPL above $329–331 into the product event.

Lower yields by themselves are not enough. The stocks need to respond.

Scenario 4: PPI/CPI come in hot

This is the week's biggest broad downside risk.

Hot inflation after a strong payroll report and a geopolitical oil shock would make a September hike much easier for the market to price aggressively.

In that environment, the most important signal is not which stock has the best AI story. It is which one refuses to break support while the discount rate rises.

Scenario 5: inflation comes in softer than feared

Then Friday's software/platform selloff can unwind quickly.

MSFT is probably the cleanest quality rebound candidate, while GOOGL and AAPL have nearby repair levels and catalysts. Chips can still outperform, but the risk/reward of chasing AMD after Friday's +4.7% becomes less attractive if money rotates back into platform megacaps.

The AI ecosystem is broadening — and so are the risks

The weekend data reinforce the core bullish argument: the physical infrastructure boom is real.

Nvidia is buying model/developer distribution through Hugging Face. Broadcom sees enormous custom-AI-chip demand through 2028. Foxconn's server business is booming. South Korea's chip exports are exploding. Dell recently raised its AI-server expectations. Meta is building custom silicon and more capacity.

But the risk surface is broadening too.

OpenAI acknowledged a recent "wiki incident" involving unintended agent misuse and said more transparency is needed around agent safety. Separately, the Seattle Times and Newsday copyright lawsuit adds another legal front around OpenAI and Microsoft.

These developments do not mean enterprise AI adoption is stopping. They do mean that the winners increasingly need to prove four things at once:

Capability.
Economics.
Safety/control.
Regulatory and licensing durability.

That is one reason I would not value every AI headline equally. A benchmark win matters less if it does not convert into deployment, revenue and defensible distribution.

Bottom line

Friday's most important signal was not that the Nasdaq fell 0.29%.

It was that semiconductors rose 3.4% anyway.

That tells us investors were still willing to pay for selected AI-hardware exposure even as hot jobs data pushed rates higher.

The weekend then made the setup more complicated:

Macro risk increased after U.S. strikes on three Iranian crude carriers.
AI-demand evidence strengthened through Foxconn and South Korean semiconductor exports.
Apple's September 9 event is now only a few days away.
PPI and CPI arrive Thursday and Friday immediately ahead of the September 15–16 Fed meeting.

So my priorities for the next 1–7 sessions are selective:

NVDA: $229–230 support; $234.75–235 breakout. Best overall combination of catalyst and tape.
AMD: strongest Friday momentum, but $478–483 is now the hard decision area rather than an easy entry.
AAPL: $318–320 support into Wednesday's event; $329–331 repair.
TSLA: $351–354 must hold before I trust a rebound; $362–365 is the first real repair.
AVGO: $360 first, then $364–367 before I call the earnings damage absorbed.
META: $610–611 support and $619–620 breakout; still the cleanest megacap trend.
MSFT: $500 is the line; $510–511 is the proof that software is catching up.
GOOGL: $337–338 support; $342–344 repair.
AMZN: $255–256 support; $261–262 then $264–266 repair.

And before I trust any of those stock levels Tuesday morning, I would look at oil, the 2-year Treasury and the 10-year Treasury.

The AI fundamental story is strong. The short-term trade depends on whether that growth can outrun a higher inflation and interest-rate premium.

Sources

Checked September 6, 2026:

Markets, rates and macro

AI and semiconductor demand

Company-specific catalysts

Price references

Market research only, not individualized financial advice. Short-term levels are observation zones based on recent trading ranges, not guarantees. Holiday gaps, geopolitical headlines, oil, inflation data, Treasury yields, regulatory news and company events can move prices directly through them.

Written and reviewed by /lico

Just writing down my thoughts, interests, and the things I learn along the way.