How One of Wall Street's Most Successful Investors Is Repositioning His AI Stock Portfolio — A 1,000%+ Surge Story

The Strategic Shift: When Billionaire Portfolio Managers Pivot

Ken Griffin, whose name resonates as a legend in hedge fund management (leading Citadel Advisors to outperform the S&P 500 by 8 percentage points over three years), recently made a telling move. In Q3, his fund divested 1.6 million Amazon shares while simultaneously adding 388,000 shares of Palantir Technologies — a company whose stock has skyrocketed 1,030% since the start of 2024. For context, even Nvidia’s remarkable 281% gain during the same window pales in comparison.

This portfolio rotation tells a story worth dissecting for anyone tracking the artificial intelligence investment thesis.

Amazon: Still a Winner Despite the Trimming

Why would a top-performing fund manager reduce exposure to Amazon? Before jumping to conclusions, understand what Amazon has actually been building.

The company operates across three major growth engines: e-commerce dominance, advertising supremacy, and cloud computing leadership through Amazon Web Services (AWS). Each segment has become a proving ground for AI applications:

E-commerce operations now feature generative AI tools powering customer service, inventory optimization, and autonomous last-mile delivery systems. The company’s Rufus shopping assistant is on track to generate approximately $10 billion in annual revenue — a category that barely existed two years ago.

The advertising business has leapfrogged to become the third-largest ad technology player globally, with AI systems that autonomously generate visual content, video, and audio assets. These tools help brands scale campaigns that previously required significant manual effort.

Cloud infrastructure through AWS continues expanding with platform services like Bedrock for AI application development, custom silicon optimized for machine learning workloads, and AI agents automating complex operational tasks.

The financial reality reflects this AI integration: third-quarter revenue reached $180 billion (up 13%), while operating margin expanded 60 basis points and operating income jumped 23% to $21.7 billion. Wall Street forecasts 18% annual earnings growth over the next three years, making the current 33x price-to-earnings multiple appear justified.

Griffin’s partial exit likely represents profit-taking rather than lost conviction — Amazon remains a top-10 holding in his portfolio.

Palantir: The Explosive Growth Story with a Valuation Warning

The flip side of Griffin’s Q3 trades involves his increased stake in Palantir Technologies, the data analytics and AI platform provider whose trajectory has been nothing short of extraordinary.

Palantir’s competitive advantage stems from ontology-based software architecture — products built around decisioning frameworks that improve continuously through machine learning. Applications range from supply chain analytics to fraud detection and battlefield intelligence.

Industry validation came from Forrester Research, which ranked Palantir as the most capable AI/ML platform available, outpacing offerings from Alphabet’s Google, Microsoft Azure, and AWS. Analysts explicitly noted Palantir is “quietly becoming one of the largest players in this market.”

The business performance matches the industry buzz. Revenue accelerated 63% to $1.1 billion in Q3 (marking nine consecutive quarters of acceleration), and non-GAAP earnings more than doubled to $0.21 per diluted share. Management credited surging demand for its AI platform capabilities.

But here lies the critical problem: Valuation has become completely detached from fundamentals.

Palantir currently trades at an extraordinary 119x sales ratio — the highest in the entire S&P 500 by a significant margin. The nearest competitor, AppLovin, sits at 45x sales. This means Palantir could lose over 60% of its value and still be the most expensive stock in the index.

Consider the mathematical reality: stock price increased 11x since January 2024, yet revenue barely doubled. This reveals the uncomfortable truth — investor enthusiasm for higher valuation multiples, rather than business fundamentals, has driven the surge. The stock traded at just 18x sales in early 2024.

Sustainability concerns are legitimate. What rises on multiple expansion alone must eventually face valuation gravity. When market sentiment shifts, the downside risk becomes substantial.

Notably, Palantir doesn’t appear in Griffin’s top 300 holdings — suggesting his Q3 purchase, while noteworthy, shouldn’t be interpreted as deep conviction in the name.

The Broader Investment Lesson

Griffin’s portfolio adjustments capture a moment when artificial intelligence dominates capital allocation decisions. Yet the disparity between Amazon (steady fundamentals, reasonable valuation) and Palantir (explosive growth, unjustifiable valuation) demonstrates why not all AI exposure carries equal risk-reward profiles.

For investors tracking the AI revolution, the key question isn’t which company has better technology — it’s which is priced for realistic expectations versus speculative exuberance.