Nvidia NVDA Bull, Base, Bear: 3-Year AI Scenario Map

Why Nvidia Needs A Scenario Map, Not A Single Price Target

Nvidia is no longer just a traditional chip designer. At roughly $ 182 per share and a multi-trillion-dollar valuation, the market is treating NVDA as the core infrastructure provider for the AI era.

The company dominates data-centre GPUs, sits at the centre of hyperscaler AI capex, and reports Q3 FY2026 earnings on 19 November, with consensus expecting about 54.8 billion dollars in revenue and strong Blackwell updates. Expectations are high, and the range of possible outcomes over the next few years is wide.

That is why a structured bull / base / bear framework is useful. It does not predict a single outcome. It clarifies what needs to be true for the current valuation to make sense, and what could break the story.

This post focuses on a 3-year horizon, roughly through late 2028, using today’s price of 182.77 dollars as the anchor.

What Actually Drives NVDA Over 3 Years

Short-term trading around earnings gets the headlines. Over three years, four bigger forces matter more than any single quarter:

1. AI data-centre capex cycle
   Whether Microsoft, Amazon, Google, Meta and peers keep spending tens of billions per year on AI infrastructure, or shift into an optimisation and “sweat the assets” phase.

2. Moat durability
   How well Nvidia defends its share and pricing power against:

   • AMD’s Instinct roadmap

   • Custom ASICs from hyperscalers (Google TPU, AWS Trainium, Microsoft Maia, etc.)

   • Other accelerators and networking vendors like Broadcom

3. Regulation and geopolitics
   Export controls into China, national security concerns, and antitrust pressure can all affect Nvidia’s addressable market and margins.

4. Product and execution quality
   The ramp of Blackwell and then Rubin, the strength of the CUDA and software ecosystem, and Nvidia’s ability to keep leading in both training and inference.

The scenarios below are built around those levers.

Bull Case: “AI Utility Wins”

• 3-year target price: about 300 dollars

• Approximate return vs today: around +64%

Story in this world

In the bull case, Nvidia ends up looking like an AI utility rather than a cyclical chip company.

• AI infrastructure capex stays elevated for longer
  Hyperscalers continue to build new AI data centres and expand existing ones. Enterprise AI adoption moves from pilots into production workflows. Agentic AI, copilots and AI-enabled automation require large, persistent compute footprints.

• Nvidia largely defends its moat
  CUDA remains the default stack for cutting-edge model training. AMD and custom chips win business, but Nvidia keeps the majority of value in premium GPUs, networking, and software. Data-centre GPU share remains very high.

• Margins stay strong
  High-end Blackwell racks and systems ship in volume at premium pricing. Networking and software add mix support. Gross margins remain in the low to mid-70 percent range.

• New legs matter, not just narrative
  Areas like automotive, edge AI, Omniverse, digital twins and AI Enterprise subscriptions grow from side stories into material contributors.

In that world, the market is willing to pay a premium multiple for high, relatively predictable earnings, because AI still looks like a long infrastructure build-out rather than a short hype cycle.

A move to around 300 dollars in three years implies roughly mid-teens annualised returns from today’s price, before any dividends.

Base Case: “Great Business, Normalised Expectations”

• 3-year target price: about 220 dollars

• Approximate return vs today: around +20%

Story in this world

The base case assumes AI is structurally important, but the growth curve gradually looks more normal.

• AI is big but not infinite
  The AI build-out continues, but hyperscalers become more disciplined on capex. They focus on utilisation, efficiency and ROI. Enterprise adoption grows, though not every workflow justifies top-end GPUs.

• Competition takes slices of the value pool
  AMD wins some notable deployments in both training and inference. Custom silicon and networking players like Broadcom capture part of the economics. Nvidia remains the leader, but dominance is less absolute than in 2023-2025.

• Margins compress, but not dramatically
  Pricing pressure, product mix and competition take a few points off gross margin. Nvidia is still very profitable, just less exceptional than at peak cycle.

• Earnings grow, multiples compress
  EPS continues to rise as revenue grows and buybacks retire shares, but the valuation multiple drifts lower as the market sees a very strong business instead of a unique hyper-growth outlier.

In this world, a move to around 220 dollars in three years would represent roughly high single-digit annualised returns from today. Nvidia remains a high-quality compounder, but the stock behaves more like a strong large-cap growth name than a speculative AI icon.

Bear Case: “Capex Hangover And Margin Squeeze”

• 3-year target price: about 120 dollars

• Approximate return vs today: around −34%

Story in this world

The bear case requires more than a slight wobble. It assumes several major risks show up together.

• AI capex rolls over faster than expected
  After multiple years of intense spending, hyperscalers shift to “doing more with what they have.” New deployments slow as CFOs focus on payback periods and utilisation metrics rather than headline GPU counts.

• Competition turns into real price pressure
  AMD’s MI series gains meaningful traction. Custom ASICs take a larger share of workloads such as search, recommendation systems, and internal LLMs. Nvidia can still sell hardware, but pricing and volume both face pressure.

• Geopolitics bites harder
  Export controls tighten further, and large parts of the Chinese AI market are effectively closed or capped. Nvidia can redirect some supply, although the overall opportunity set shrinks.

• Valuation re-rates sharply
  Growth slows, margins compress, and the market stops treating Nvidia as a special case. Valuation multiples move closer to those of other prominent semiconductor names with solid, but not explosive, growth.

Under this combination, a move toward 120 dollars over three years is consistent with slower revenue growth, lower margins and a lower earnings multiple. That would mean negative double-digit annualised returns from today’s price.

Framing The Risk/Reward From Here

Using today’s price of 182.77 dollars as the anchor, the simplified three-year scenario map looks like this:

These are not precise forecasts. They are signposts that link narrative to numbers:

• If AI infrastructure spending stays intense for longer and Nvidia keeps a dominant position, the bull world becomes more likely.

• If AI becomes an important but more ordinary growth driver, with competition and capex discipline both rising, the base path fits better.

• If hyperscalers pivot to optimisation, competition bites harder and geopolitics tighten, the bear set-up starts to describe reality.

The probabilities will shift as new information lands, particularly around:

• Guidance and commentary from the Q3 FY2026 earnings call

• Evidence on hyperscaler and sovereign AI capex plans

• Gross margin trends as Blackwell ramps and competition responds

• Any change in export controls or regulatory pressure

Final Thoughts and Regulatory Note

Nvidia sits at the centre of the current AI cycle. The share price already reflects that. A scenario framework like this does not settle the debate over whether NVDA is “cheap” or “expensive”. It simply makes the trade-offs clearer: how much of the future is already priced in, and what sort of world is implied by different price paths.

This overview is based on public information and broad assumptions as of 18 November 2025. It does not account for individual objectives, financial situations or tax considerations.