To The Moon

Wall Street has quickly moved from correction fears back to record-high excitement.

On Monday, the Dow gained 1.32% to close at a record 53,178. The S&P 500 rose 1.48% to 7,600, while the Nasdaq climbed 2.13%. Falling oil prices and lower Treasury yields gave growth stocks another boost, while strong earnings brought capital back into technology and AI names.

That leaves investors facing an uncomfortable question:

Are earnings pulling the market higher, or is fear of missing out creating a high-valuation melt-up?

Tonight may provide an important answer.

$Advanced Micro Devices(AMD)$ and $Arista Networks(ANET)$ both report after the closing bell. AMD represents the computing layer of the AI buildout. Arista represents the networking layer connecting those computing clusters.

If both companies deliver strong results and raise guidance, the rally gains broader fundamental support.

If expectations prove too aggressive, the market may discover how little room for disappointment remains at current valuations.

Why are the indexes moving higher again?

The first driver is the macro environment.

Crude oil fell around 5% on Monday as investors reacted to signs of possible de-escalation between the United States and Iran. Lower oil prices helped pull Treasury yields down, reducing near-term pressure on expensive growth stocks. (Reuters)

The second driver is corporate earnings.

Among the 304 S&P 500 companies that had reported through Friday, earnings were growing 29.3% year over year, while 85.2% had beaten analyst expectations. Those are strong enough numbers to support the broader market narrative that profits are still catching up with prices. (Reuters)

Market breadth also improved during Monday’s rally. Advancing stocks outnumbered decliners by 2.62 to 1 on the New York Stock Exchange and by 3.01 to 1 on the Nasdaq. Trading volume was also above its recent average. (Reuters)

So this rally is not being driven entirely by speculation.

The concern is that expectations are now extremely demanding.

Is this becoming a melt-up?

A melt-up occurs when investors rush into the market because they fear being left behind, pushing prices higher faster than earnings and fundamentals can comfortably justify.

Several melt-up characteristics are beginning to appear:

• Indexes are returning to record territory very quickly

• AI and mega-cap technology stocks are attracting concentrated flows again

• Small improvements in oil, yields or earnings are producing outsized gains

• Investors are increasingly buying before earnings rather than waiting for confirmation

The market can continue rising under these conditions.

The risk is that the reward for a normal earnings beat becomes smaller, while the punishment for a miss becomes much larger.

That creates a difficult setup.

Strong companies may still rise, yet investors chasing them immediately before earnings could face double-digit gaps if guidance fails to clear elevated expectations.

At this stage, the key question is no longer whether a company can beat consensus.

It is:

How far above consensus must it deliver before investors are satisfied?

Tonight’s AI test begins with AMD

AMD will release its second-quarter results after the U.S. market closes, followed by a management conference call at 5:00 p.m. Eastern Time. (Advanced Micro Devices, Inc.)

The company previously guided for approximately $11.2 billion in second-quarter revenue, plus or minus $300 million, representing roughly 46% year-over-year growth at the midpoint. AMD also guided for a non-GAAP gross margin of approximately 56%. (Advanced Micro Devices, Inc.)

The headline revenue number will matter.

The composition of that revenue will matter more.

Can EPYC keep taking server share?

AMD’s data-center growth still relies heavily on EPYC server processors.

AI data centers require more than accelerators. They also need CPUs for data preparation, workload scheduling, storage management and general-purpose computing.

That means AMD needs to demonstrate that EPYC demand remains strong even as Intel attempts to rebuild momentum in the server market.

A strong EPYC quarter would show that AMD has two working engines:

• Server CPUs generating current revenue

• AI accelerators creating the next phase of growth

A slowdown would raise questions about whether Intel’s recovery is beginning to limit AMD’s share gains.

When do MI450 and Helios become meaningful revenue?

Investors already know that AMD is building a larger AI accelerator roadmap.

Tonight, they need clearer evidence on execution:

• When will MI450 begin contributing meaningful revenue?

• How quickly can Helios rack-scale systems enter production?

• Are customer deployments expanding beyond initial commitments?

• Will third-quarter and fourth-quarter guidance reflect faster AI shipments?

AMD needs to move beyond being viewed as an alternative supplier.

The stronger long-term case requires customers to adopt AMD as a second full-scale AI computing platform alongside Nvidia.

Can margins rise with AI revenue?

Revenue growth alone may not be enough.

Advanced packaging, high-bandwidth memory, systems development and early production costs can weigh on profitability as new AI products scale.

If revenue beats while gross margin or forward guidance disappoints, investors may conclude that AMD is sacrificing too much profit to win market share.

The strongest result would combine:

• Data-center revenue above expectations

• Strong EPYC demand

• Faster AI accelerator shipments

• Stable or improving gross margin

• A meaningful increase in forward guidance

AMD is fighting on three fronts

AMD is often presented as the main alternative to $NVIDIA(NVDA)$ in AI accelerators.

Its competitive challenge is broader than that.

In server CPUs, it must continue gaining share from $Intel(INTC)$.

In cloud infrastructure, it must also defend the x86 ecosystem as $Arm Holdings(ARM)$ architecture becomes more important in custom chips designed by hyperscalers.

That creates a three-way contest:

Nvidia: dominant AI accelerator platform
Intel: attempting a server CPU recovery
Arm: expanding through energy-efficient, customized cloud processors

AMD’s advantage is its ability to offer CPUs, GPUs, networking and software as part of a wider data-center portfolio.

Its challenge is that every part of that portfolio now faces serious competition.

Tonight’s earnings need to prove that AMD can challenge Nvidia without losing momentum against Intel and Arm.

Arista answers the second half of the AI question

Arista Networks will also release second-quarter results after the closing bell, with its management call scheduled for 4:30 p.m. Eastern Time. (Arista Networks)

Arista reported first-quarter revenue of $2.709 billion, up 35.1% year over year. Its non-GAAP operating margin reached 47.8%, and non-GAAP earnings per share were $0.87. (Arista Networks)

For the second quarter, management guided for:

• Revenue of approximately $2.8 billion

• Non-GAAP operating margin of 46%–47%

• Non-GAAP earnings per share of approximately $0.88 (Arista Networks)

Those numbers set a high bar.

Arista must now show that AI infrastructure spending continues to move beyond GPUs and into the networks connecting them.

Does the AI boom still require more networking?

Large AI clusters depend on thousands of accelerators communicating with one another.

Without enough bandwidth and low-latency connectivity, expensive GPUs cannot operate at their full potential.

This makes Arista a key indicator for the next stage of the infrastructure cycle.

Investors will be watching demand for:

• High-speed Ethernet switching

• 800-gigabit and next-generation networking

• AI spine architectures

• Optical and scale-up connectivity

• Cloud data-center expansion

Arista has also introduced products aimed at denser and more efficient AI infrastructure, including networking systems designed to reduce rack usage and floor-space requirements. (Arista Networks)

If its guidance rises again, it would suggest that hyperscaler spending is spreading across the full AI stack.

Customer concentration remains a major issue

Arista’s largest cloud customers can generate enormous orders.

That creates strong growth during expansion cycles, but it also makes results sensitive to the purchasing schedules of a limited number of companies.

The market will want clarity on whether cloud customers such as Microsoft and Meta are:

• Increasing networking budgets

• Accelerating AI-cluster deployments

• Moving to newer Ethernet architectures

• Digesting equipment purchased in previous quarters

• Developing more internal alternatives

Strong AI demand with stable customer concentration would support the growth story.

Any sign of order delays or spending pauses could trigger concerns that networking investment is becoming less synchronized with GPU purchases.

Can Arista protect its exceptional margins?

Arista’s valuation is supported by both growth and unusually high profitability.

Its first-quarter non-GAAP operating margin was 47.8%, while management guided for 46%–47% in the second quarter. (Arista Networks)

That means investors will judge the company on two standards at once:

Revenue must continue growing rapidly.

Margins must remain close to best-in-class levels.

If Arista wins additional AI networking orders by accepting lower pricing, revenue may outperform while the stock still reacts negatively to margin pressure.

A clean earnings beat would require strong sales, resilient margins and another constructive outlook.

Why AMD and Arista must be viewed together

These companies sit at two critical points in the same capital-spending chain.

AMD measures compute demand.

Its results show whether customers are buying more server CPUs and AI accelerators.

Arista measures network demand.

Its results show whether those customers are also building the high-speed infrastructure required to connect larger clusters.

If both companies beat expectations and raise guidance, the message would be powerful:

AI investment is moving through the entire data-center stack.

That could support related names across several groups.

Compute and semiconductors:
$NVIDIA(NVDA)$, $Broadcom(AVGO)$, $Taiwan Semiconductor(TSM)$

Networking and connectivity:
$Credo Technology(CRDO)$, $Marvell Technology(MRVL)$

Memory and storage:
$Micron Technology(MU)$, $SanDisk(SNDK)$, $Western Digital(WDC)$

Power and cooling:
$Vertiv(VRT)$, $Eaton(ETN)$

If AMD performs well while Arista disappoints, the market may conclude that compute demand remains strong but networking purchases are becoming less consistent.

If Arista performs well while AMD disappoints, the infrastructure cycle may still be healthy while AMD faces company-specific execution or competitive problems.

If both disappoint, investors may begin questioning whether AI spending has already been priced too aggressively across the supply chain.

So, should investors chase the market here?

The indexes can continue setting new highs.

Strong earnings, lower oil prices and improving market breadth provide real support.

The risk comes from treating every technology stock as equally attractive simply because the index is rising.

Three signals matter most from here.

Follow earnings revisions, not only index levels

High valuations can remain high when forward earnings estimates keep rising.

Companies that increase guidance deserve more attention than those merely beating a low quarterly estimate.

Watch whether the rally broadens

A healthier AI rally should extend beyond Microsoft, Amazon and Nvidia.

Semiconductors, networking, storage, power and software need to participate.

Record indexes alongside weakening industry breadth would suggest a more fragile, concentrated advance.

Watch the reaction after earnings

In an expensive market, price reaction can be more revealing than the reported numbers.

A company that beats and still falls may have entered earnings with expectations that were impossible to satisfy.

A company that reports an ordinary quarter and holds its gains may have already absorbed much of the valuation pressure.

TigerComments Take

This rally has enough earnings support that it cannot be dismissed as pure speculation.

At the same time, record index levels have reduced the market’s tolerance for ordinary results.

The earlier stage of the AI trade rewarded companies for having exposure to the theme.

This stage is rewarding companies that can:

• Convert demand into revenue

• Raise forward guidance

• Protect margins

• Generate cash while continuing to invest

AMD and Arista now face the same market test from different sides.

AMD must prove that AI computing can develop into a multi-platform market rather than remaining dominated by Nvidia.

Arista must prove that hyperscalers are still expanding the networks surrounding those accelerators.

If both pass, the index rally gains a broader foundation.

If the results are mixed, capital will continue concentrating in the small group of companies delivering the clearest orders, margins and cash flow.

The takeaway:

Record highs can still be bought—but chasing the index out of FOMO is very different from buying companies with rising earnings expectations.

What happens after tonight’s AI earnings?

A. AMD and Arista both deliver, and the AI rally broadens
B. Core leaders keep rising while the rest of the market stays divided
C. Valuations have priced in too much, triggering profit-taking
D. Wait for the reports and buy only after the market reacts

Are you adding AI exposure before the results, or waiting to see how much good news is already priced in?

Disclaimer: This post is for market discussion only and does not constitute investment advice. AMD and Arista are scheduled to report after the U.S. market closes on August 4, 2026. Actual results and guidance should be confirmed through company filings.