The Earnings Illusion: How Corporate Buybacks Are Masking a Massive S&P 500 Vulnerability

It’s worth taking a moment to consider a number that the majority of people outside the financial sector are unaware of. American businesses spent $165.63 billion repurchasing their own stock in just July of 2025. It’s not a quarter. That’s just one month. It was 88% higher than the previous all-time peak set in July 2007, which, for those who recall, was the summer before the financial crisis started to unravel, according to Birinyi Associates. As they say, history rhymes, and this particular rhyme is becoming more pronounced.

The number for the entire year is even more bizarre. S&P 500 companies had already spent $926.1 billion on share repurchases by the end of 2025, which was about $108 billion more than the previous year-to-date record. In the face of tariffs, an Iranian conflict, an unpredictable White House, and a Fed that is unsure whether inflation is under control or not, corporate America decided that repurchasing its own shares at record highs was the best use of its cash. This might be a sign of true confidence. It’s also possible that this is the largest bet-on-yourself event in the history of the modern market—and not in a good way.

This is the subtle tactic that no one discusses at CNBC’s opening bell. Consider a business with 100 outstanding shares that makes $100 in profit. Earnings per share: $1. Let’s say the company makes exactly the same $100 the following year, with no new products, margin expansion, or real growth, but it repurchases 20 shares. In a flash, earnings per share reach $1.25. A “growth” figure of 25% appears on the screens. Analysts give a nod of approval. Models are modified by portfolio managers. On Robinhood, retail investors see green. In reality, nothing changed within the organization. The denominator was the only variable that changed. That is the earnings illusion, and it is currently operating on an industrial scale.

Leading the way are the financial, technology, and communication services sectors, which together repurchased $689 billion of their own stock in 2025. With only $55 billion, utilities have been particularly cautious. That split is a story unto itself. The businesses that are most vulnerable to interest rate sensitivity and AI hype are repurchasing the most. Almost charmingly, those with real long-term infrastructure responsibilities are investing in their own plants and grids.

Like a ghost that won’t go away, the Hewlett-Packard saga continues to linger in the background of this discussion. HP spent roughly $67 billion on buybacks between Carly Fiorina, Mark Hurd, and Leo Apotheker—more than the company made in those same years. In quarterly reports, it looked fantastic. Executive Comp was energized by it. For a moment, it raised the stock.

The company eventually had to split in two and eliminate 80,000 jobs because it had no more blockbuster products or a compelling software and services story. HP “destroyed itself by downsizing its labor force and distributing its profits to shareholders,” according to William Lazonick, the UMass-Lowell economist who has likely written more about this than anyone alive. It was sobering to watch it happen in real time. It is quite different to watch the rest of the S&P 500 replicate the strategy.

The optimist’s counterargument, which merits a fair hearing, is that buybacks are simply a flexible method of returning capital, that businesses with nothing better to do with cash are making a sensible decision, and that the calculations benefit long-term shareholders. Sometimes it’s true. However, the mid-2010s Reuters analysis, which examined data from 3,297 publicly traded companies, revealed an unsettling finding: among companies that reported R&D and repurchased their own shares, the percentage of net income allocated to innovation had decreased from more than 60% in the 1990s to less than 50% for the majority of the 2010s. To understand what’s going on, you don’t need a PhD. Spending on the secondary market does not result in the creation of new manufacturing facilities, goods, or employment.

Additionally, Oxfam outlined the distributional angle in October 2025, pointing out that big businesses are paying far more in shareholder payouts than they do in federal taxes. It’s not against the law. Depending on your political beliefs, it might not even be incorrect. However, this raises an unsettling question about the S&P 500’s true nature in 2026: is it a reflection of actual corporate productivity or is it a carefully crafted financial tool meant to continue rising until it is unable to?

It’s difficult to ignore the fact that the market was only a year away from losing half of its value when buyback activity last reached such a high peak in the middle of 2007. That does not imply that it will recur. It does imply that the earnings season that appears to be the cleanest in years may not be as clean as the headlines portray. The mirror has a lovely appearance. It is worthwhile to inquire about what lies beneath.