NVIDIA is one of the main beneficiaries of the growth of artificial intelligence. The company produces graphics processing units, networking equipment, and software solutions used to train and run AI models. But is the market price of its shares justified?

Let’s try to value NVIDIA using Peter Lynch’s method with Valestor.

Fair value

At the time of review, NVIDIA shares (NASDAQ: NVDA) were trading at $186.64. The estimated fair value according to Peter Lynch’s model is $198.70.

So the model points to a potential upside of about 6%. By the formal result, the stock is undervalued.

However, the margin of safety is relatively small. This is important to keep in mind: even if the company continues to grow rapidly, the current price already reflects highly optimistic expectations from investors.

Company type and earnings growth

NVIDIA belongs to the category of fast-growing technology companies. The main driver of its development is demand for computing power for artificial intelligence, data centers, and high-performance computing.

The company’s net income grew at the following rates:

• Over 5 years — 65% per year.
• Over 3 years — 59% per year.
• Over 1 year — 28%.

The dynamics of net income are also impressive: from $4.3 billion in 2020 to $120.1 billion in 2025.

At the same time, growth rates are gradually slowing. This does not necessarily mean deterioration in the business, but it does show how difficult it is to maintain the previous pace after such rapid scaling.

P/E and PEG

For valuation using Lynch’s method, PEG is especially important — the ratio of stock price to earnings per share adjusted for the rate of earnings growth.

PEG is 0.94. A value below 1 is usually considered a sign of potential undervaluation.

But there is an important nuance: NVIDIA’s actual five-year net income CAGR is 65%, while for the PEG and fair value calculations, Valestor caps the growth rate at 25%.

This is not accidental: Peter Lynch believed that growth above 25% per year is hard to sustain over the long term.

Fair price according to the Lynch formula

The fair price is calculated from the following data:

TTM EPS — $7.95.
Limited earnings growth rate — 25%.
The calculation looks like this:
25 × $7.95 = $198.70.
This is the value the model uses as the fair value of one NVIDIA share.

Competitive advantage and economic moat

NVIDIA’s main advantage is not only the performance of its graphics processors. The company has created an entire ecosystem around CUDA, a software platform for parallel computing.

Developers and companies have spent years building software for this architecture. As a result, switching to alternative solutions can require significant costs and time.

Another advantage is the comprehensive data center solutions: accelerators, networking equipment, and software. NVIDIA sells not just individual chips, but an entire computing platform.

This creates a serious competitive barrier, although it does not make the company invulnerable.

Main risks

The main risk is market expectations. Investors are counting on continued large-scale investment in AI infrastructure. If spending by major technology companies slows down, this could affect demand and NVIDIA’s valuation. The market reacted very nervously to the victory of the Democrats in the congressional elections and expects new restrictive AI laws.

It is also worth considering competition from AMD, specialized AI chip developers, and the in-house processors of major cloud companies.

Another factor is restrictions on exporting advanced chips, dependence on manufacturing partners, and a high concentration of demand among the largest customers.

Conclusion

NVIDIA demonstrates exceptional earnings growth and has a strong competitive advantage thanks to the CUDA ecosystem and comprehensive AI solutions.

Using Peter Lynch’s model, Valestor estimates the fair value of the stock at $198.70 versus a market price of $186.64. The upside is around 6.46%, which is relatively modest.

It is important to remember that the Lynch formula is based on historical earnings growth and does not predict future results. For NVIDIA, this is especially significant: maintaining high growth rates is not guaranteed, and market expectations are highly volatile.

For that reason, the model result should be treated as a starting point for a very cautious analysis.