INVESTING
riod in which underwriting relationships can distort an analyst’ s incentives.
The authors tested the analyst bias idea by looking at stocks with heavy short interest. Here, the prediction market’ s advantage was significantly smaller, which suggested short sellers had already forced more honest analyst behavior.
This means investors who treat the analyst consensus as an unbiased probability assessment of earnings outcomes are working with flawed information. 2. Continuous Information Updating While analyst consensus is only periodically refreshed and can become stale, prediction market prices update continuously, incorporating new signals as they arrive.
The researchers found that a crowd captures information that an analyst consensus has not yet processed. 3. The Wisdom of Crowds This is the paper’ s most novel contribution. Unlike previous work, the paper turns to blockchain transaction data to study crowd forecasting. What they’ ve found is consistent with observations by James Surowiecki in his book The Wisdom of Crowds, which finds that groups making decisions with an aggregation of data come up with better outcomes than people working by themselves.
The prediction market, the paper said,“ aggregates information from a broader and more diverse set of participants than the relatively homogeneous sell-side analyst community. Participants may include supply-chain observers, employees, customers, informed retail traders, quantitative funds, and institutional investors who do not publish research. No individual analyst has access to this full set of dispersed signals.” 4. An Idiosyncratic Information Advantage This explains how prediction markets add the most value.
When a company’ s earnings move in lockstep with its broader industry, analysts can rely on peer announcements and publicly available sector data. Prediction markets have little to add here. But
While Wall Street analyst consensus is only periodically refreshed and can become stale, prediction market prices update continuously.
when earnings are idiosyncratic— driven by factors specific to a company that are known by dispersed stakeholders like employees, customers, and supply-chain participants( and before they are known by analysts)— the prediction markets retain a substantial advantage. The crowd’ s edge is largest precisely where the crowd has an information advantage that traditional analysis cannot replicate.
The Mechanism Matters
This paper provides some of the cleanest evidence to date that decentralized information aggregation can outperform what’ s collected by credentialed professionals— not because the crowd is smarter on average, but because it channels diverse, independent signals that no centralized expert can access. Hayek’ s 1945 insight was right: The knowledge of the world is dispersed, and prices do the best job of bringing it all together.
Another takeaway from the paper is that an analyst’ s bias is not just random noise— it’ s something predictable and exploitable. Understanding where analysts are most compromised helps investors know when to discount the consensus most heavily.
Also, continuous updating of information matters. Real-time price formation captures value that snapshot forecasts miss.
For investors, that means taking consensus-based earnings surprises with a grain of salt. Are they based on real-time information or do they happen when analysts are correcting for a bias?
A Note Of Caution
The authors are candid about the limitations of their work. The sample covers only six months and is limited to firms with active prediction markets— likely larger, more liquid companies with sufficient public interest to attract prediction market participants. The mechanisms are inferred from observed patterns rather than directly measured. And while insider trading was ruled out as a driver, the authors acknowledge it cannot be fully excluded.
These are real limitations. The evidence is preliminary and the sample is short. It will be essential to replicate these results over longer horizons, and broader samples will be needed before we can draw strong conclusions.
What this paper demonstrates, however, is that the presumed superiority of credentialed experts is not guaranteed— especially when those experts operate within institutional structures that systematically distort their incentives. The market for information, like the market for goods, may be more efficient than any centralized bureau of forecasters.
LARRY SWEDROE is the author or co-author of 18 books on investing, including his latest Enrich Your Future. He is also a consultant to RIAs as an educator on investment strategies. This article is for informational and educational purposes only and should not be construed as specific investment, accounting, legal, or tax advice.
44 | FINANCIAL ADVISOR MAGAZINE | SEPTEMBER / OCTOBER 2026 WWW. FA-MAG. COM