
Over the last two years, artificial intelligence has been a dominant theme driving global equity markets. Companies involved in designing and manufacturing the chips that power Artificial Intelligence (AI) systems have delivered exceptional returns, with investors rushing to gain exposure to what many believe is the next technological revolution.
Recent weakness in semiconductor shares has sparked an important debate among investors: are we seeing the start of a “chip wreck,” or simply a pause after an extraordinary run?
To answer that question, it is important to separate the AI story from the investment story.
The AI story remains incredibly compelling. Technology giants continue to spend billions of dollars on data centres, cloud infrastructure and advanced computing capabilities. Demand for the most sophisticated chips remains strong, and many semiconductor companies are still reporting impressive earnings growth.
The investment story, however, is becoming more complicated. Markets are forward-looking. Investors are not buying shares based on what companies earned last quarter; they are investing based on what they believe those companies will earn years into the future. The challenge is that expectations for the semiconductor sector have become exceptionally high.
Many of the leading chip companies are now being valued on the assumption that AI spending will continue growing at an aggressive pace for years to come. While that may ultimately prove correct, it leaves very little room for disappointment.
This is where the concept of a “chip wreck” begins to emerge.
The biggest risk that may be facing the sector is not that AI adoption suddenly slows or disappears, but rather that growth starts to normalise. If large technology companies begin moderating their AI spending, or if investors become concerned that the return on those investments is lower than expected, semiconductor shares could face meaningful pressure even if company earnings remain strong.
History provides several examples of this phenomenon. Markets often move through cycles where a genuine long-term trend becomes so widely accepted that investors begin assuming the good news will continue indefinitely. Eventually, expectations become difficult to exceed, and even strong results can disappoint investors.
Another factor worth considering is the sheer amount of capital currently being invested into AI infrastructure. According to Goldman Sachs[1], between $600 billion and $800 billion will be spent on AI infrastructure in 2026. Technology companies are spending hundreds of billions of dollars to secure computing power and data centre capacity. The four largest spenders (Amazon, Microsoft, Alphabet and Meta) are expected to spend approximately $725 billion on AI-related activities in 2026. This increase is significant when compared to the $416 billion spent in 2025 and $251 billion in 2024. AI spend in 2026 will almost be three times what was spent in 2024.¹
While demand remains strong today, there is always the possibility that supply eventually catches up. If that happens, pricing power and profit margins across parts of the semiconductor industry could come under pressure.
Investors should also be aware of the concentration risk that has developed in global equity markets. A relatively small number of technology and semiconductor companies have been responsible for a significant portion of market returns over the past two years. If you look at the MSCI All Country World Index (ACWI), the index has a 32% exposure to the Information Technology sector. When leadership becomes this concentrated, markets can become more vulnerable to periods of volatility if sentiment shifts.
That does not mean investors should abandon the sector. In fact, the long-term outlook for semiconductors remains attractive. Artificial intelligence, automation, cloud computing, autonomous vehicles and digital infrastructure are all likely to require increasing amounts of computing power for many years to come. The world is becoming more dependent on technology, and technology is becoming more dependent on advanced semiconductors.
The key question for investors is therefore not whether semiconductor companies will continue to grow, but whether current valuations already reflect much of that future growth. Great companies do not always make great investments at every price. Even when the underlying theme remains intact, periods of excessive optimism can create short-term risks for investors.
In closing
A full-scale “chip wreck” is unlikely. The underlying demand drivers supporting the sector are simply too powerful. However, after such a significant rally, investors should expect increased volatility and potentially more selective performance within the sector.
The next phase of the AI trade is likely to be driven less by excitement and more by execution. Companies will need to continue delivering earnings growth, maintaining margins and proving that the enormous investment in AI infrastructure can generate sustainable returns.
The AI revolution is real. The demand for chips is real. But as markets have reminded investors many times before, even the strongest themes can experience turbulence when expectations begin running ahead of reality.
For long-term investors, diversification remains critical. The semiconductor sector may continue to play an important role in portfolio growth, but successful investing has never been about relying on a single trend. It is about building portfolios that can benefit from opportunities while remaining resilient when markets inevitably become more uncertain.
__________________________
Sources:
- Statista “Big Tech’s AI Spending to Reach $725 Billion in 2026”; published 30 April 2026.
- Yahoo Finance “Meta, Microsoft, Amazon, and Alphabet are about to spend a shocking amount of money to dominate the AI era”; published 3 June 2026.
- Goldman Sachs “Why AI Companies May Invest More than $500 Billion in 2026”; published 18 December 2025
[1] Source: Goldman Sachs “Tracking Trillions: The Assumptions Shaping the Scale of the AI Build-Out” published 1 May 2026.