Inside Prediction Markets: How Sharps Outsmart Wall Street

In the high-stakes world of prediction markets, a new breed of trader is quietly reshaping how the world forecasts everything from election outcomes to military strikes. Known as “sharps,” these sophisticated traders have made millions on platforms like Polymarket and Kalshi, wagering on outcomes that range from geopolitical conflicts to Rotten Tomatoes scores. As The New York Times reports in a deep-dive feature, these markets are challenging the very foundations of traditional Wall Street forecasting. According to Wikipedia, Polymarket is a cryptocurrency-based prediction market founded in 2020 that allows users to bet on outcomes ranging from sports matches to military conflicts.

The Explosive Growth of Prediction Markets

The numbers are staggering. Combined monthly volume on Polymarket and Kalshi surged from $1.8 billion in April 2025 to $24.2 billion in April 2026 — a more than thirteen-fold increase in just one year. By the end of 2025, monthly volume had already reached $13 billion. These platforms are now valued at $9 billion (Polymarket) and $11 billion (Kalshi), with Intercontinental Exchange (ICE) investing $2 billion in Polymarket in October 2025 at an $8 billion valuation.

This explosive growth has transformed prediction markets from a niche curiosity into a potentially systemic financial infrastructure. But who is actually making money in these markets?

The Rise of the Sharps

The answer, according to extensive research, is a tiny minority. On Polymarket, just 0.1% of accounts — roughly 2,000 traders — capture 67% of all profits, generating approximately $500 million. Meanwhile, more than 70% of users lose money. As Michael Boss, a former professional poker player who now executes 60 trades per minute on Kalshi and has earned over $668,000, bluntly put it: ordinary users “simply have no chance of winning.”

These sharps employ sophisticated strategies that would be familiar to any Wall Street quant: real-time data feeds, AI algorithms, high-frequency trading, and statistical modeling. Jonathan Stoll-Ryan, a student whose firm ranks in the top five on Kalshi, spends over $200,000 per year on data and AI infrastructure alone. The anonymous trader known as “Domer” has made $2.6 million on Polymarket since January 2022, leveraging skills honed as an online poker professional.

Outperforming Wall Street Analysts

The most striking finding from the research is that prediction markets are demonstrably better at forecasting than traditional financial analysts. According to research by Dr. Roberto Gomez-Cram of London Business School, prediction markets correctly forecast 78% of earnings outcomes, compared to just 62% for Wall Street analysts. “The romantic idea of crowd wisdom may be overstated,” Dr. Gomez-Cram noted, “but the future still looks promising.”

However, the mechanism behind this accuracy is not the “wisdom of crowds” that prediction market advocates often cite. Instead, research shows that approximately 3% of accounts drive price discovery on these platforms. The accuracy comes from an informed minority, not the collective judgment of the masses.

The 2024 Election: A Watershed Moment

Prediction markets gained mainstream attention during the 2024 U.S. presidential election. On election eve, Polymarket and Kalshi gave Donald Trump approximately 55% odds of winning, while traditional polls showed Kamala Harris ahead. Trump won, and the markets’ accuracy earned them newfound credibility.

But the election also raised troubling questions. A single French trader wagered approximately $30 million on Trump across four accounts, ultimately winning $85 million. While Polymarket investigated and found no evidence of market manipulation, the incident highlighted how a small number of deep-pocketed bettors can move markets. Polymarket CEO Shayne Coplan has defended the platform, calling prediction markets “the most accurate thing we have as mankind right now.”

Insider Trading and Regulatory Concerns

The darker side of prediction markets has emerged through a series of insider trading scandals. In January 2026, a newly created Polymarket account netted over $400,000 from bets on Nicolás Maduro’s capture hours before the U.S. military operation was publicly announced. A U.S. Special Forces soldier was later arrested and charged with leaking classified information.

Perhaps most alarming was the case of the Israeli Air Force. Personnel were indicted for betting on the timing of strikes on Iran using classified information, with one officer earning $244,000. During his interrogation, a crewmember stated that “the entire squadron is on Polymarket, the entire air force is betting.” According to the Anti-Corruption Data Collective, 52% of military long-shot bets succeed, compared to 14% of all bets — a statistical anomaly that strongly suggests insider information is being exploited.

The Regulatory Landscape Shifts

The regulatory environment has undergone a dramatic transformation. Under the Biden administration, the CFTC fined Polymarket $1.4 million in 2022 and the FBI raided founder Shayne Coplan’s home in November 2024. But under the Trump administration, enforcement has eased considerably. The CFTC and DOJ ended their probes into Polymarket in July 2025, and Donald Trump Jr. now serves as an advisor to both Polymarket and Kalshi, with his firm 1789 Capital investing in the platform.

This has raised conflict-of-interest concerns. Yale professor Jeffrey Sonnenfeld warned that “given the conflicted relationship of the First Family, CFTC oversight could be compromised.” The CFTC has just one-eighth the staffing of the SEC, yet oversees billions in weekly trading volume. Enforcement remains largely reactive rather than preventive.

The Accuracy Paradox and What’s Next

Prediction markets present a paradox: they are remarkably accurate, but their accuracy is driven by a tiny informed minority — and in some cases, by individuals with access to non-public information. This raises fundamental questions about whether these markets are genuinely “smarter” than traditional forecasting, or whether they are simply vehicles for exploiting information asymmetries.

Polymarket has continued to expand, launching earnings markets in September 2025 and private-company markets in May 2026. The platform has also faced bans in France, Italy, Singapore, Switzerland, Poland, Indonesia, and other countries concerned about gambling and national security implications.

As these markets grow — with combined volumes now exceeding $24 billion monthly — the question is no longer whether prediction markets will influence finance and politics, but how regulators, investors, and the public will respond to a system where a handful of sharps hold most of the cards, and where the line between informed trading and insider trading grows increasingly blurred.

This article is based on reporting from The New York Times Magazine (May 26, 2026), with additional context from Wikipedia and academic research cited in the original feature.