The Nasdaq-100 index is home to 100 of the largest non-financial companies listed on the Nasdaq stock exchange, including each of the "Magnificent Seven" stocks, which have a combined value of $14.7 trillion:
- Apple: $3 trillion.
- Microsoft: $2.9 trillion.
- Nvidia (NVDA 4.11%): $2.7 trillion.
- Amazon: $2 trillion.
- Alphabet (GOOG 1.52%) (GOOGL 1.70%): $1.9 trillion.
- Meta Platforms: $1.4 trillion.
- Tesla: $811 billion.
Their size is just one of the reasons Wall Street recognized this group as the Magnificent Seven in 2023. Each of the seven stocks has a tendency to outperform the broader market over the long term, and the underlying companies are solidified leaders in their respective industries.
On April 2, President Donald Trump announced plans to impose tariffs on nearly all physical products imported into America, which sent investors flocking to safe-haven assets like cash on fears of a global economic slowdown. As a result, the Nasdaq-100 is in the throes of a bear market after plunging by as much as 23% from its February record high.
But that might be a big opportunity for investors. Shares of Nvidia and Alphabet are down by more than 20% from their all-time highs, yet most of the products they sell are either exempt from Trump's tariffs or were never impacted to begin with. Here's why it might be a great idea for investors to buy the dip, with the intention of holding Nvidia and Alphabet for the long run.
Image source: Nvidia.
The case for Nvidia
Nvidia sells the world's most advanced graphics processing units (GPUs) for data centers, which set the benchmark when it comes to artificial intelligence (AI) development. It all started with the H100 GPU, which helped the company win 98% of the market in 2023, as it was the best chip for training large language models (LLMs) and performing AI inference. These models generated "one-shot" responses to user inputs, which was useful at the time, but they often made mistakes.
Rather than feeding LLMs an endless amount of training data, developers are now building "reasoning" models that spend more time thinking before generating a response. In other words, they make better use of the data they already have and weed out inaccuracies before rendering a response for the user. Some of the best reasoning models available today include OpenAI's GPT-4o series, Anthropic's Claude 3.7 Sonnet, DeepSeek's R1, and Alphabet's Gemini 2.5.
NASDAQ: NVDA
Key Data Points
According to Nvidia CEO Jensen Huang, these models require a staggering 100 times more computing power than traditional LLMs. He says each response they generate consumes 10 times more tokens (words and symbols) because of how much time they spend thinking. Plus, that makes them much slower to render responses, so GPUs need to be 10 times faster to offset the delays -- otherwise, users might grow tired of the whole experience.
As a result, Nvidia's Hopper architecture (which is at the foundation of the H100 GPU) is no longer sufficient. Instead, developers are transitioning to Nvidia's Blackwell-based GPUs like the GB200, which can perform AI inference at 30 times the speed of the H100. The company also revealed an upgraded architecture called Blackwell Ultra during the GTC 2025 conference in March. Blackwell Ultra delivers up to 50 times more performance than Hopper.
Nvidia's data center business generated a record $115.2 billion in total revenue during its fiscal year 2025 (ended January 26), which was a whopping 114% increase from the prior year. But Huang believes AI data center infrastructure spending will top $1 trillion annually by 2028 as reasoning models grow more power-hungry, so Nvidia's opportunity is set to grow significantly from here.
Investors can buy Nvidia stock at a very attractive valuation right now. It trades at a price-to-earnings (P/E) ratio of 37.7 as of this writing, which is a 36% discount to its 10-year average of 59.6:
NVDA PE Ratio data by YCharts
Semiconductors are exempt -- for now -- from Trump's tariffs on Taiwan, which is where many of Nvidia's chips are fabricated. However, the company does face uncertainty because it relies on hundreds of billions of dollars in AI data center spending from just a handful of customers to fuel its growth, so if trade tensions dent the global economy, some of those buyers could revise their capex plans.
However, this is likely to be a short-term headwind because countries are already negotiating trade deals with the Trump administration. Therefore, investors who buy Nvidia stock at the current price could still do extremely well over a time frame of three to five years.
Image source: Alphabet.
The case for Alphabet
Alphabet is the tech conglomerate behind Google, YouTube, self-driving car developer Waymo, and more. Since Google has been the window to the internet for over two decades, Alphabet has a bigger stockpile of valuable data with which to develop powerful AI models than almost any other company, and it's putting it to good use.
Alphabet launched the Gemini 2.5 Pro reasoning model earlier this month, and the company says it outperforms the latest models from OpenAI, Anthropic, DeepSeek, and even Elon Musk's xAI across several benchmarks. It powers a stand-alone chatbot called Gemini, but it will also help Alphabet further improve the new AI features in Google Search -- and this is critical because the advertising dollars brought in by the search engine still account for over half of the conglomerate's total revenue.
NASDAQ: GOOGL
Key Data Points
Alphabet introduced AI Overviews for Google Search last year, which combines text, images, and links to third-party websites to produce AI-generated responses to almost any query. Overviews appear at the top of the traditional search results, so they can save users from sifting through web pages to find the information they need. So far, Alphabet says Overviews monetize just as well as the traditional search format, so it appears they won't disrupt the company's golden goose.
Alphabet doesn't just develop AI for its own purposes -- it also sells a range of services to help businesses and developers achieve their goals in this space via the Google Cloud platform. It offers access to powerful data center infrastructure and ready-made models like Gemini to help developers accelerate their AI software projects. In the final quarter of 2024, Google Cloud customers were using eight times more computing capacity than they were 18 months earlier, which highlights how quickly AI adoption is scaling.
Alphabet is currently the cheapest stock in the Magnificent Seven. Based on the company's $8.04 in earnings per share last year, its stock trades at a P/E ratio of just 19.5:
TSLA PE Ratio data by YCharts
The rest of the Magnificent Seven stocks trade at an average P/E ratio of 31.3 (excluding Tesla because its sky-high valuation is an outlier), so Alphabet stock would have to soar by 60% just to catch up.
Alphabet's advertising revenue could take a hit if global trade tensions spark an economic slowdown because businesses would probably trim their marketing budgets. But aside from that indirect consequence, the company primarily sells digital products and services that are excluded from tariffs at the moment.
Setting trade tensions aside, Alphabet's leadership position in the AI race could make it a fantastic addition to any portfolio over the long term, especially at the current price.
