The S&P 500 (^GSPC 0.01%) delivered back-to-back annual gains of over 25% in 2023 and 2024 (when including dividends). The only other time the index had such a strong two-year run was during the dot-com internet boom in 1997 and 1998. As was the case back then, stocks in the technology and tech-adjacent industries drove most of the upside this time around.
The "Magnificent Seven" is a group of seven companies that lead different segments of the tech space. They earned the nickname for their enormous size and their tendency to outperform the broader S&P 500. In fact, investors who didn't own them during 2023 and 2024 likely underperformed the index by a very wide margin:
Data by YCharts.
However, the S&P 500 is off to a bumpy start to 2025 due to global trade tensions, which were sparked by President Donald Trump's "Liberation Day" tariffs. All of the Magnificent Seven stocks are trading down from their all-time highs, presenting long-term investors with a compelling opportunity. Shares of Meta Platforms (META 3.05%) and Amazon (AMZN 0.81%) are down 9% and 15%, respectively, from all-time highs hit in early 2025 (as of this writing).
Here's why investors might want to buy the dips.
Image source: Getty Images.
The case for Meta Platforms
Meta is the parent company of social networks like Facebook, Instagram, and WhatsApp, which serve over 3.4 billion people every single day. Acquiring new users is becoming harder because nearly half the planet is already using those platforms, so the company is trying to boost engagement instead. The longer each user spends online, the more ads they see, and the more money Meta makes.
Meta uses artificial intelligence (AI) in its recommendation engine to learn what type of content Facebook and Instagram users enjoy viewing, and it uses that information to show them more of it. During the first quarter of 2025, CEO Mark Zuckerberg said this strategy led to a 6% increase in the amount of time users were spending on Instagram over the last six months, and a 7% increase for Facebook.
But Meta is also using AI to help businesses craft better ads. Soon, Zuckerberg says a business will simply have to tell Meta its goals (like brand awareness or selling a specific product) and its budget, and an AI assistant will handle the rest -- that includes designing the creative (the text, image, or video) and defining the audience. Most small businesses don't have their own marketing teams, so this could be a powerful tool that boosts Meta's share of the digital advertising market.
Meta AI is another innovation from the social media giant. It has become one of the most popular chatbots in the world with almost 1 billion monthly active users, despite only launching last year. It's built on Meta's Llama family of large language models (LLMs), which are now among the most intelligent in the AI industry. Meta plans to spend up to $72 billion on data center infrastructure and chips this year, most of which will be geared toward advancing Llama even further and meeting inference demand for Meta AI.
Meta has generated $25.64 in earnings per share (EPS) over the last four quarters, so based on its current stock price of around $650, it trades at a price-to-earnings (P/E) ratio of 25.2. That makes it the second-cheapest Magnificent Seven stock ahead of only Alphabet, which is battling a series of regulatory issues:
Data by YCharts.
Because of Meta's progress in AI, its solid (and growing) profits, and its valuation, I think the company is on its way to the exclusive $2 trillion club.
The case for Amazon
E-commerce remains the single biggest contributor to Amazon's overall revenue, but investors are currently more focused on the Amazon Web Services (AWS) cloud computing platform. It offers hundreds of solutions to help businesses navigate the digital world, but it's also trying to dominate the three core layers of AI: hardware infrastructure, LLMs, and software.
AWS operates data centers filled with chips from leading suppliers like Nvidia, and it rents the computing capacity to AI developers for a profit. But Amazon also designed its own chips, and the new Trainium2 can save developers up to 40% on AI training costs compared to competing hardware.
AWS also developed a family of LLMs called Nova, which includes the new NovaSonic speech-to-speech model for conversational AI applications. Ready-made models eliminate the need for developers to create their own from scratch, so using them can significantly accelerate their AI software projects. The Nova family is available on Amazon Bedrock, along with other LLMs from leading third parties like Meta and Anthropic.
On the software front, AWS now offers an integrated virtual assistant called Q, which can write computer code to help developers create software more quickly. It can also analyze internal data to provide businesses with actionable insights. In addition, Amazon developed a separate virtual assistant for Amazon.com called Rufus, which helps shoppers compare products and make more informed decisions.
Amazon generated $155.6 billion in total revenue during the first quarter of 2025. AWS accounted for just 19% of that total ($29.2 billion), yet it was responsible for 63% of the entire company's operating income. Segments like e-commerce operate on razor-thin margins, so AWS is the profit engine behind the whole organization, which is why investors monitor it so closely.
AWS has become a leader in AI infrastructure and services, and I think that will be Amazon's golden ticket to the ultra-exclusive $3 trillion club in the next couple of years, where it could sit alongside Microsoft, Apple, and Nvidia.
