Equity Prediction Markets
Profit-fueled corporate cynicism
Leonardo Barcellos & Scott Emett
Journal of Accounting and Economics, forthcoming
Abstract:
A large literature documents that many households underinvest in equities, but the beliefs underlying this behavior remain poorly understood. We study Americans' cynicism toward public companies as a belief structure that helps explain limited market participation, examining its prevalence, underlying mechanisms, and strategies to attenuate it. In a nationally representative survey, we find that corporate cynicism is widespread and negatively predicts individuals' propensity to invest in stocks. Building on these findings, we conduct a series of experiments and provide causal evidence that corporate cynicism reduces stock investments. We further show that perceiving companies as profit-maximizing fuels corporate cynicism, and that cynicism can be reduced by reshaping beliefs about profit maximization or profit itself. Specifically, cynicism declines when companies highlight broader stakeholder responsibilities in narrative disclosures, when people are prompted to take an investor perspective, and when financial media focus on how corporate profits benefit most Americans versus deepen inequality.
Human Edge, Machine Limits: Algorithmic Herding and AI-Human Competition
Winston Wei Dou et al.
University of Pennsylvania Working Paper, March 2026
Abstract:
We develop a theoretical framework of competition between AI-powered and human investors with heterogeneous sophistication. Human investors possess superior private information but are constrained by bounded rationality, modeled through a cognitive-hierarchy structure. AI investors, by contrast, learn through reinforcement learning and autonomously optimize trading over time. Despite starting from heterogeneous algorithms, AI investors endogenously converge to similar trading rules, generating algorithmic herding. We show that the most rational human traders can nevertheless outperform even AI investors, because AI learns from data generated by average, not frontier, human behavior. AI profitability is limited by three forces: the private-information advantage of humans, the price-stabilizing trading of the most sophisticated human investors, and the rising price impact of AI trading as its market share grows through algorithmic herding. Together, these mechanisms reveal the limits of algorithmic superiority in financial markets.
The Present Value of Future Market Power
Thummim Cho et al.
Review of Financial Studies, forthcoming
Abstract:
We introduce a present-value identity relating a firm's market value to expected future markups, output growth, discount rates, and investments. Distinguishing current from expected markups reveals five empirical facts: (1) Expected markups account for half the rise in U.S. firm values since 1980. (2) The rise in aggregate expected markups reflects market-share reallocation toward high-expected-markup firms and within-firm increases. (3) Expected markups are linked to intangible investments. (4) They relate negatively to discount rates over time but (5) positively to abnormal returns across firms. Finally, variation in long-term expected markups is primarily associated with asset prices rather than current markups.
Interactivity and Illusions of Ability: How Using Generative AI Affects Investor Judgments
Joe Croom
Journal of Accounting Research, May 2026, Pages 681-719
Abstract:
I use the setting of generative AI (GenAI) to examine how processing tool interactivity affects investors' self-assessments of ability and willingness to invest. Although GenAI can help investors process financial information, I theorize that the interactive nature of GenAI blurs the boundaries between investors' own abilities and those of GenAI, prompting investors to discount their reliance on GenAI and misattribute its abilities to themselves. I rely on the advantages of a laboratory setting to disentangle the interactive element of GenAI from the mere presence of GenAI assistance. Across three experiments, I find that the interactivity underpinning GenAI heightens investors' self-assessments of their own abilities and increases their willingness to invest, despite this interactivity not improving, and in fact hindering, their actual processing of information provided by GenAI. My study thus highlights one potential cost of using GenAI and other highly interactive processing tools.
The Trillion Dollar Bonus of Private Capital Fund Managers
Ludovic Phalippou
Management Science, forthcoming
Abstract:
Carried interest (carry) is the main performance-based component of compensation for private capital fund managers. Using fund-level cash flows and fee terms for more than 12,000 funds, we estimate which funds are in the carry and the total amount earned. Aggregate carry exceeds one trillion dollars and accounts for 18% of investor profits, about equal the contractual value-weighted rate of 19%. The difference reflects the role of hurdle rates and the relatively smooth distribution of fund outcomes. Carry is strongly related to both performance and fund size, and past carry is a stronger predictor of future performance than past returns.
A Comparison of Agentic AI Systems and Human Economists
Serafin Grundl
Federal Reserve Working Paper, April 2026
Abstract:
This paper compares agentic AI systems and human economists performing the same causal inference tasks. AI systems and humans generally obtain similar median causal effect estimates. While there is substantial dispersion of estimates across model instances, the human distributions of estimates have wider tails. Using AI models as reviewers to compare and rank "submissions," the following ranking emerges regardless of reviewer model: (1) Codex GPT-5.4, (2) Codex GPT-5.3-Codex, (3) Claude Code Opus 4.6, and (4) Human Researchers. These findings suggest that agentic AI systems will allow us to scale empirical research in economics.
Automation and High-Skill Labor Markets: Evidence from Robo-Advisors
Matthew James Flynn & Ishitha Kumar
Texas State University Working Paper, April 2026
Abstract:
Does artificial intelligence substitute or complement high-skill labor? We study financial robo-advisors and their impact on the labor market for human advisors. Using novel data on robo-advisors and advisor employment, we find that robo-advisors increase, rather than reduce, human advisor employment. To address endogenous adoption, we exploit cross-state regulatory entry costs interacted with a national financing shock affecting technology firms. States experiencing greater robo-advisor entry exhibit meaningful growth in advisor employment, particularly among firms serving retail investors. Investor-level evidence shows robo-advisor exposure increases subsequent engagement with human advisors, consistent with market expansion and complementarity between automation and professional financial services.
Who Clears the Market When Passive Investors Trade?
Marco Sammon & John Shim
Review of Financial Studies, forthcoming
Abstract:
We find that firms are the primary sellers who clear the market for index fund buying, providing shares at a nearly one-for-one rate. Most demand-side institutions trade in the same direction as index funds rather than accommodating passive demand. We use two instruments for index fund demand and show that firms causally respond to exogenous passive demand, with prices serving as the coordinating mechanism. Firms satisfy passive demand mostly through nonprimary market issuance, for example, through employee stock-based compensation. Our results suggest that passive investing has systematically supplied capital to firms by enabling equity issuance over the last two decades.