After one of the semiconductor sector’s most painful selloffs in six years, $NVIDIA(NVDA)$ , $Advanced Micro Devices(AMD)$ and other chip stocks staged a strong recovery on Friday. The rebound has raised an important question: Was this week’s violent AI selloff a temporary correction, or are investors finally reaching the end of the AI trade?
I think it is too early to declare the bottom, but there are reasons to believe the outlook is not as negative as Thursday’s market reaction suggested.
1. The AI demand story has not disappeared
The biggest distinction between this correction and previous technology selloffs is that AI infrastructure demand remains substantial.
Hyperscalers continue investing heavily in data centres, advanced processors, networking equipment and power infrastructure. Nvidia remains central to AI computing, while AMD is competing for a larger share of the accelerator market.
The important point is that companies are not suddenly abandoning AI investment. The market is questioning how much they are spending, how quickly those investments will generate returns and whether current valuations already reflect too much future growth.
That is a valuation problem, not necessarily a demand problem.
2. Investors are becoming more selective about AI winners
The AI trade has become crowded, with investors previously rewarding almost any company associated with artificial intelligence.
That environment is changing.
Companies with strong revenue growth, improving margins, competitive advantages and visible customer demand may continue attracting capital. However, businesses relying heavily on future expectations without demonstrating commercial returns could face further pressure.
Nvidia and AMD are particularly interesting because their products are directly connected to AI infrastructure spending. However, even strong businesses are not immune to valuation corrections.
3. Two factors could support a recovery
There are two developments worth watching.
First, AI adoption continues expanding beyond technology companies. Businesses are increasingly exploring AI applications to improve productivity, automate processes and reduce operating costs. If these applications translate into measurable financial benefits, the current infrastructure investment cycle could prove more sustainable.
Second, easing concerns about data-centre development and infrastructure restrictions could support the next phase of investment. AI expansion requires not only chips but also electricity, cooling systems, networking equipment and physical facilities.
This creates opportunities across a broader group of companies rather than just the largest semiconductor names.
4. Why another selloff remains possible
Despite Friday’s rebound, several risks remain.
• Valuations: Some AI-related stocks still trade at substantial premiums based on earnings expectations that may take years to materialise.
• Capital expenditure: Microsoft, Amazon, Alphabet and Meta are committing enormous amounts of money to AI infrastructure. Investors increasingly want evidence that these investments will generate attractive returns.
• Competition: AMD and other chipmakers are challenging Nvidia, while major technology companies are also developing their own AI chips.
• Market positioning: When investors become heavily concentrated in the same popular trades, even relatively small changes in sentiment can trigger sharp corrections.
A strong rebound after a major selloff can also reflect short covering and bargain hunting rather than a genuine improvement in market fundamentals.
5. What I would watch next
For Nvidia (NVDA), I would focus on continued data-centre revenue growth, margins and evidence that customers are maintaining their spending commitments.
For AMD (AMD), the key question is whether it can convert growing demand for AI accelerators into meaningful market share and earnings growth.
For $Broadcom(AVGO)$ , custom AI chips and networking remain important areas to monitor.
I would also watch the earnings and capital expenditure guidance of the major hyperscalers. Their spending decisions could ultimately determine whether the semiconductor recovery has room to run.
My conclusion: The AI selloff may be losing momentum, but the next rally needs fundamental support.
I don’t think the long-term AI investment opportunity has disappeared. However, the market is moving into a more demanding phase where revenue growth, profitability and returns on capital matter more than excitement about AI alone.
Friday’s rebound is encouraging, but I would prefer to see several sessions of sustained strength, improving market breadth and continued support from earnings before calling this a confirmed bottom.
The next opportunity may not necessarily be buying every AI stock that has fallen. It may be identifying the companies whose earnings can genuinely justify their valuations.
For discussion purposes only, not financial advice
Comments