With Amazon (NASDAQ: AMZN) shares down 11% since the start of the year, a lot of investors are wondering if it’s time to buy the dip. Like many tech stocks, the diversified e-commerce giant has been hit by a souring macroeconomic outlook and possible generative artificial intelligence (AI) fatigue. Let’s explore what the next 12 months could have in store.
Amazon is a massive $2.04 trillion company and it isn’t done growing yet. Fourth-quarter net sales jumped 10% year over year to $187.8 billion. And while that may seem like a small growth rate, we are talking about hundreds of billions here in annual sales. Its recent top-line growth was mainly powered by a strong showing from its North American e-commerce unit, which had a particularly successful holiday shopping season.
Amazon’s bottom line also continues to flourish in the post-pandemic economy. It has been about four years since CEO Andrew Jassy took the helm and shifted away from the growth-at-all-costs strategy of his predecessor, Jeff Bezos, to prioritize profitability. Jassy’s approach helped Amazon expand operating income by 61% year over year to $21.2 billion in Q4, with the cloud computing division, AWS, contributing around half of the total.
Management is not yet finished cutting costs. On the heels of other large layoffs, Amazon intends to eliminate close to 14,000 managerial positions this year, which analysts at Morgan Stanley predict could save it between $2.1 billion and $3.6 billion annually.
Amazon isn’t cutting costs from a position of weakness. It also enjoys long-term growth drivers like AI. The company plays a middleman role in the cloud infrastructure side of the opportunity, essentially accumulating hardware from chipmakers like Nvidia to power its massive data centers, and renting out this processing power to the software companies creating AI applications.
Management plans to expand Amazon’s role in ways that will somewhat encroach on Nvidia’s territory. This month, it announced plans to offer discounted access to servers powered by its custom-made AI chip, Trainium, which is designed to train generative AI large language models (LLMs).
According to online news outlet The Information, these chips match the performance of Nvidia’s H100 GPUs at about a quarter of the cost. And while Nvidia has moved on from its Hopper (H) architecture to its more advanced Blackwell-based chips, Amazon still looks poised to capture a share of the AI chip market.
AI-related demand helped power the growth of AWS in the fourth quarter. This segment is vital for Amazon because it boasts significantly higher margins than e-commerce.
