Prediction markets and the future of Covid-19

Forecasters have used all sorts of tools to gauge where the U.S. and the world are headed with Covid-19, ranging from epidemiologic models to historical ones. Here’s another possibility: prediction markets.

Prediction markets use a wisdom of crowds approach to aggregate beliefs and predict future outcomes. They typically function through bets in futures contracts that pay out on expiration of defined “yes or no” outcomes. The Iowa Electronic Markets, for example, successfully predicted the outcomes of every presidential election from 1988 to 2000 to within 1.5 percentage points.

Prediction markets have their own agnostic efficiency and logic. The investor goal is to make money. The market-maker goal is to accurately predict outcomes without the “noise” of morality, science, ideology, politics, or culture. The wisdom of crowdsourced betting is harnessed to make accurate predictions in complex scenarios with significant uncertainties.

In 2008, online prediction market Intrade predicted that Barack Obama would win 364 electoral college votes. He won 365.

The prediction market platform might be used for pandemic forecasting — and may be more effective than traditional approaches.

Morality, science, ideology, politics, and culture are the very forces that have shaped personal and national policy responses to the Covid-19 pandemic. As we approach three years of life with SARS-CoV-2, the virus that causes Covid-19, and deliberations move to the “living with the virus” phase, decision-making alternatives are needed.

Science offers much legitimate guidance about risk levels and actions to mitigate them — whether and when to wear masks, test, quarantine, get vaccinated or boosted, and the like — though each of these have been hotly and sometimes acrimoniously debated. This is where prediction markets could earn their stripes.

Applying a detached prediction market lens approach may produce provocative — and perhaps more accurate — pandemic forecasts than pure evidence-driven approaches. As a thought experiment, let me explore how smart money can be deployed in a virtual Covid-19 prediction marketplace betting on the “winners” and “losers” in the next year of Covid. In real life, this tool has been only sporadically applied to Covid-19 and has not fulfilled its potential as input to the Covid zeitgeist.

Investment analysts typically make their forecasts by integrating three factors: the fundamentals (where the pandemic is headed); key societal trends (the external environment); and potential demand for products and services (revenue models).

Fundamentals. In this bettors’ scenario, 2023 will feature significantly decreased serious illness and deaths, largely as a result of durable steady-state population immunity through both ongoing vaccination and repeated cycles of natural infection and re-infection. Elderly, debilitated, and immune-deficient individuals will continue to bear the brunt of the virus. Neither new variants nor dramatic changes in prevention or treatment will significantly modify this scenario.

Societal trends. These include growing emphasis on protecting vulnerable individuals and populations, growing acceptance of the risk of exposure among non-vulnerable individuals, and more “leading from behind” relaxation of official regulations and guidelines in response to an increasingly risk-tolerant population. There will be significant ongoing loosening of restrictions like quarantine, isolation, masking, and testing, as well as waning interest in and uptake of vaccination.

Revenue models. Due to congressional resistance, federal funding for Covid-19 relief will dramatically decline. This fall we may see the last round of free vaccines, tests, and treatments. Some of the slack will be taken up by the commercial marketplace. Consumer out-of-pocket costs will increase and the uninsured and indigent will have diminished access to these products. Overall, commercial demand for goods and services in the Covid space will decrease, while some selected areas, like rapid testing, may see an increase, including through direct-to-consumer marketing.

With that background, here are my “market-maker” predicted Covid-19 trends for 2023:

Bet on these (long bets in investor terminology)

Long Covid. This is an ongoing American health catastrophe that will draw significant and increasing investment and revenue for basic and clinical research, health care, disability expenses and advocacy.

Protecting the vulnerable. There will be increasing individual and societal focus on safeguards for this group. Even as the rest of society relaxes, vulnerable individuals and populations will require increased vaccines and treatments, non-pharmaceutical interventions, and social support.

Pandemic preparedness. Much remains to be learned and applied in preparation for the spillover of the next animal virus to humans. While federal funding is uncertain, foundation and philanthropic engagement are likely to be high.

In-person activities. Schools, colleges, sports, entertainment, retail, and travel are not only likely to return to normal, but will overshoot. Office-based employment won’t get back to what it was, but will increase considerably from pandemic lows.

New point-of-care diagnostics. Consumers have become familiar and comfortable using at-home rapid tests. While the demand for conventional antigen tests will continue to decrease, there will be a need for new diagnostics that can differentiate various causes of cold and flu-like upper repository symptoms. Late-stage development is already occurring in this area.

Neutral

Oral antivirals. Paxlovid and molnupiravir are both effective in preventing serious Covid-19 among those over age 65, but not in younger age groups. Expect new Medicare coverage but decreased utilization in the under-65 population.

Bet against these (short bets in investor terminology)

Covid protocols. Most schools, colleges, workplaces, and sports leagues will be abandoning these.

Non-pharmaceutical interventions (masking, ventilation, quarantine, isolation). Going forward, demand will be limited mainly to the vulnerable and the worried-well populations.

Vaccines and testing. The incentives for these fade as the perceived threat diminishes.

U.S. funding for the global pandemic. With decreasing domestic concern about Covid-19 and stretched foreign aid budgets and higher priorities such as supporting Ukraine, this will fall by the wayside.

The CDC. The announced internal reorganization of the Centers for Disease Control and Prevention does not eliminate external issues of independence, interagency and state competition and prerogatives, lack of coordination among key players, and politicization of public health. The recapture of credibility and trust will prove elusive.

This is an exercise, not advice

A reminder to potential bettors: This is an exercise intended to capture insights into where the pandemic is going, and one I hope will nudge people to put their conceptual money where their mouths are. This illustration of trends is intended to provide fresh insights for policymakers, the public, and the media that are frequently not evident through more conventional forecasting tools.

For those eager to place real-life bets, diligent research and the stock market may offer validation of sagacity. There is a fine line, however, between pandemic prediction and gambling. Caveat emptor.

Steven Phillips is a medical epidemiologist and the vice president for science and strategy for the Covid Collaborative.