Diego Vallarino, PhD - Ideas

Polymarket Shows You a Probability. It Does Not Show You Whether That Probability Deserves to Be Believed.In my first post on Polymarket, I showed that its prices should not be interpreted as hundreds of independent forecasts. Once contracts are required to describe economically related states of the world, most apparent cross-market cascades disappear.

Polymarket Is Not One Market, and Its Prices Do Not Correct at the Same SpeedIn my previous post, I argued that Polymarket’s prices should not be interpreted as independent forecasts. Using a database of 14,646,957 trades, 1,249,016 wallets, 464 comparable markets, and 1,009 major price-regime events, I showed that most apparent cross-market cascades disappear once contracts

Polymarket’s Prices Are Not Independent Forecasts, and That Changes What They MeanThe study began with a database containing 500 Polymarket markets, 14,646,957 trades, 1,249,016 distinct wallets, approximately 3.9 million wallet-market observations, and more than 51,000 market-day observations.

Nonlinear Geoeconomic Cascades and Spectral Instability in Global Economic Network.1. Global interdependence as a nonlinear system.

A Quantitative Political Economy in a Geopolitically Fragmented World1. From Market Movements to Political Signals At first glance, the joint evolution of the Dow Jones, the S&P 500, Tesla, gold, and the U.

The Real Strategic Risk Is Not What You Get Wrong, It’s What You Keep Talking About.Over the last year, I’ve sat in dozens of meetings in Washington, DC. More recently, I had similar conversations in New York, taking advantage of the holidays to reconnect with friends across finance, technology, and policy.

AI Adoption: A Strategic Responsibility for Boards in Reorganizing Work.Last week, I had the opportunity to meet with colleagues and business leaders in New York and Philadelphia. Despite differences in sector and company size, a common theme emerged in every conversation: artificial intelligence is no longer peripheral—it is now central to corporate strategy.

This Halloween, I Fear Linearity More Than Monsters: Why GAT Beats DiD for Modern CausalityEvery year around Halloween, we enjoy the thrill of imaginary fears—ghosts, monsters, shadows moving in the dark. But as a data scientist and economist, my real nightmare is different.

Can Data Alone Reduce Poverty in Emerging Economies?Introduction: The Promise and the Question The debate about poverty reduction often revolves around familiar tools: social transfers, targeted subsidies, or labor market reforms. Yet, a quieter revolution is underway—one that asks whether data itself can act as a redistributive asset.

Why Firms Fail Even When Management Gets It Right.How Much Firm Survival Heterogeneity Is Structurally Mediated? Boards are trained to look inward when performance deteriorates or risk materializes. The standard diagnostic sequence is familiar: Was the strategy flawed? Did management execute poorly? Were incentives misaligned? Did leadership undere

The day I uncovered structural alpha from Open Banking networksLet economic intuition guide the machine learning. And insist that any alpha story be as statistically rigorous as it is economically coherent.

Second-Best AI: Structural Constraints and Institutional Opportunities in Latin America.1. Situation Despite the growing global prominence of artificial intelligence (AI), Latin America exhibits a highly unequal and fragmented insertion into this new technological frontier.

Page updated

Google Sites

Report abuse