Technology stocks can fall quickly when high expectations meet signs of weaker demand. Recent selling highlights one statistic worth watching: much of the artificial intelligence economy may depend on only a few customers. That concentration does not guarantee a crash. It does, however, create a risk investors should examine before assuming AI growth will continue at its recent pace.
I am Taylor Sohns, CEO of LifeGoal Wealth Advisors, a Certified Investment Management Analyst, and a Certified Financial Planner. My work includes helping oversee more than $500 million. From that perspective, the goal is not to predict every market decline. It is to identify risks that prices may not fully reflect.
The Statistic That Could Matter Most
Nvidia has become a central supplier for the AI spending cycle. At the time of this analysis, it was the largest public company by market value. Its revenue had also grown about 100% from the prior year.
That growth rate is exceptional for a company of Nvidia’s size. It helps explain the enthusiasm surrounding its shares and the wider technology sector.
Yet headline growth doesn’t tell the whole story. Nvidia’s regulatory filing showed that five customers accounted for roughly 70% of revenue. Those buyers are commonly linked to major cloud and data-center operators, often called hyperscalers.
“Just five customers accounted for 70% of that revenue.”
This figure changes how investors should view the growth. Revenue can rise rapidly while still resting on a narrow group of buyers. If even one large customer reduces spending, the effect could be meaningful.
Customer concentration is not proof that demand is weak. Large technology suppliers often rely on a limited number of major buyers. The concern is that Nvidia’s valuation may assume those buyers will keep spending at an unusually high rate.
View this post on Instagram
How AI Spending Flows Through the Market
The AI investment chain starts with companies building models and services. These firms need massive computing capacity. Cloud providers respond by purchasing advanced chips, servers, networking equipment, and data-center space.
Nvidia sits near the center of that chain because its graphics processors are widely used to train and operate AI models. Hyperscalers purchase those systems, then rent computing capacity to AI developers and business customers.
The spending chain can be summarized in four steps:
- AI developers seek computing capacity to train and run models.
- Cloud providers build data centers to meet expected demand.
- Chipmakers and equipment suppliers receive large orders.
- Investors value those suppliers based on continued growth.
Each step depends partly on the one before it. If end users pay enough for AI services, developers can purchase more computing capacity. Cloud companies may then continue ordering chips and equipment.
The risk appears when demand is concentrated near the start of the chain. According to the figures discussed here, two companies, OpenAI and Anthropic, accounted for about 70% of hyperscalers’ AI-related revenue.
That suggests a large section of the AI spending cycle may depend on a small number of private AI developers. The market may look broad because many public companies participate. However, the source of demand may be much narrower.
Why Concentration Can Increase Market Risk
Concentrated revenue creates several possible pressure points. A major customer could delay an order, negotiate lower prices, or develop its own hardware. It could also struggle to raise capital.
AI companies spend heavily on chips, cloud services, staff, electricity, and model development. Their ability to maintain that spending depends on funding and commercial revenue.
If revenue doesn’t catch up with costs, outside capital may become more important. Higher interest rates, weaker investor demand, or disappointing product adoption could make that capital harder to obtain.
A slowdown could then travel through the chain. AI developers might purchase less computing capacity. Hyperscalers could reduce capital spending. Suppliers could face slower orders and lower growth forecasts.
Stock prices may react before reported revenue falls. Markets often move when investors revise future expectations, not only when current results weaken.
This matters because high-valued technology shares usually require strong future growth. Even a healthy company can see a large stock decline if results miss optimistic forecasts.
What OpenAI’s IPO Timing May Signal
OpenAI’s possible initial public offering has attracted attention because public markets would provide more financial visibility. Required filings would let investors review revenue, expenses, cash flow, customer trends, and capital needs.
Reports of delays or uncertainty around an IPO can raise questions. They are not, by themselves, evidence of financial weakness. A private company may postpone a listing for many reasons, including market conditions, governance, regulation, or strategy.
Still, the lack of public financial statements limits what outside investors can verify. That matters if OpenAI represents a large share of demand supporting public cloud and chip companies.
“This whole tech stack depends on two companies.”
The statement is intentionally direct, but the underlying lesson is more measured. Investors should know how much reported growth depends on a few private firms whose finances receive less public scrutiny.
Why a Technology Crash Is Not Certain
I am not saying technology stocks must crash. Concentration is a warning sign, not a market forecast.
Several factors could support continued growth. AI use may spread across software, health care, finance, manufacturing, research, and consumer services. More businesses may purchase computing capacity as useful applications emerge.
Hyperscalers also operate established businesses outside AI. Their cloud platforms, advertising services, software products, and consumer operations can help fund data-center investment.
Nvidia may also gain from continued technical leadership and strong demand for its systems. A concentrated customer base can remain profitable if those customers are financially sound and keep expanding.
The central issue is not whether AI has value. The issue is whether current stock prices already assume years of rapid, steady growth.
A promising technology can produce lasting economic gains while some related stocks still fall. The internet created major businesses, yet many technology shares suffered severe losses when expectations ran ahead of reality.
How Investors Can Assess the Risk
Investors do not need to abandon technology shares because of one statistic. They should compare potential returns with the risks created by high valuations and concentrated demand.
Useful questions include:
- What percentage of revenue comes from the largest customers?
- Are major buyers increasing or reducing capital spending?
- Does growth depend on a few privately funded AI companies?
- How much future growth is already reflected in the stock price?
- Could the portfolio withstand a sharp decline in technology?
Company filings can provide more useful information than earnings headlines. Investors should review customer concentration, accounts receivable, capital spending, inventory, and management’s guidance.
Changes in those figures may offer early clues. Rising inventory could signal slowing demand. Larger receivables may indicate customers are taking longer to pay. Lower spending guidance could affect suppliers before revenue declines appear.
Diversification also matters. Owning a market index may still leave an investor heavily exposed to a small number of large technology companies. Index weights should be reviewed alongside individual stock positions.
A portfolio should match an investor’s time horizon and loss tolerance. Money needed soon generally should not depend on highly volatile shares. Long-term investors may tolerate more movement, but they still need reasonable position sizes.
Focus on Risk Instead of Predictions
Market commentary often presents two extreme choices. Either AI will support endless growth, or the sector is about to collapse. Neither conclusion is supported by one data point.
A better approach is to recognize both opportunity and risk. AI investment may remain strong, while customer concentration makes the sector more sensitive to changes at a few firms.
Investors should also separate business quality from stock price. A company can report impressive results and still be a poor investment at an excessive valuation. Another company can face short-term problems yet offer value at a lower price.
Recent weakness in technology stocks does not prove that a larger decline has started. It does remind us that popular investments can become fragile when expectations rise faster than verified demand.
The 70% concentration figures deserve attention because they reveal how connected the AI trade may be. If spending continues, suppliers could keep growing. If one or two major buyers pull back, the effects could spread quickly.
My recommendation is simple: read the filings, study who pays the bills, and avoid treating rapid growth as guaranteed. Investors do not need certainty to make sound decisions. They need a clear view of the risks they are accepting.
A technology crash is not inevitable. Yet if one occurs, investors may look back and wonder why they overlooked it.
Frequently Asked Questions
Q: Does customer concentration mean Nvidia’s revenue will fall?
No. It means a small group of customers has an outsized effect on results. Revenue could keep growing if those customers maintain their spending. The risk rises if one or more buyers reduce orders.
Q: Why do OpenAI and Anthropic matter to public technology stocks?
Both companies use large amounts of computing capacity. Their demand can support cloud providers, chipmakers, and data-center suppliers. If their spending changes, public companies tied to that demand may feel the effect, too.
Q: Should investors sell technology stocks because of this risk?
Not based on this statistic alone. Investors should review valuations, diversification, customer trends, and personal time horizons. Position sizes should reflect the possibility of sharp price swings.
Image Credit: Pexels
