Jun 25, 2018

Stories from a Real World Monetary Policy Experiment

Patrick Horan Program Manager, Program on Monetary Policy, Chad Reese Managing Editor

Just over a year ago, Scott Sumner and David Beckworth, economists with the Mercatus Center’s Program on Monetary Policy, launched a creative experiment aimed at improving how we think about monetary policy.

Sumner and Beckworth collaborated with Hypermind, a UK-based prediction market, to create their own futures market that allowed others to predict the nominal gross domestic product (NGPD) growth rate for the United States. They chose NGDP growth because they believe it is a better indication of macroeconomic stability than the inflation rate.

While participants did not invest any real money in the experimental market, the most successful traders were given Amazon gift certificates equal to what would be their dollar-denominated earnings. Of the 362 participants, 224 made a virtual profit, including a top winner who earned over $1,000 in Amazon gift certificates.

The project just completed its first year successfully, and we spoke with some of the top traders to get their views on the project.

Like in any real market, participants varied widely in their motivations and strategies.

Mehdi Kheredine, for example, bet on strong growth predictions from the first quarter all the way through the end, a strategy that looked risky at first, but paid off eventually. “It was right after September that I was able to raise myself to first place. Before that, I was in the bottom of the ranking with a negative score.”

Another top participant, Justin Irving, was confident in his econometrics background, and relied heavily on forecasting models as the basis of his predictions. Justin noted that he largely ignored public policy changes during the course of the year because “markets take all of that into account.”

Still others, like Tony Henke, placed among the top finishers despite avoiding NGDP growth estimates entirely. Henke “just noticed a certain bias across several Hypermind markets that made me think it was somewhat more likely than not that the NGDP market was too pessimistic.”

We noticed a few overlapping qualities among the winners we interviewed. Virtually all said they are interested in economics, and many of them said they became interested in the subject at an early age. Many are fans of Sumner’s blog, The Money Illusion, and several are regular participants in Hypermind markets. Only a few pointed to policy changes such as the recent tax reform law as something that impacted their predictions, and almost nobody saw the appointment of new Federal Reserve Chairman Jerome Powell as having a large impact. This may imply that participants see Powell’s approach to monetary policy as largely being a continuation of former Chair Janet Yellen’s approach.

The value of prediction markets comes from their ability to channel the “wisdom of crowds” by taking into account a diversity of viewpoints and approaches. Some participants approach a market with a particular view that seems counterintuitive at first, but profit by holding on for the long game. Others use sophisticated models to predict what will happen, and still others use a more behavioral approach, teasing out what they view as errors among other market participants. As market signals change and as market participants learn more information, they update their priors and change their views as they deem necessary. The opportunity for profit and threat of loss also encourages participants to act as rationally as possible and not be beholden to ideology or other concerns.

The second year of the prediction market experiment is ongoing. If you’re interested in helping us learn more about the ability of markets to improve monetary policy as well as earning real prizes, you can sign up here.

Support Mercatus

Your support allows us to continue bridging the gap between academic ideas and real-world policy solutions.Donate