Forward the original title ’Don’t overestimate the efficiency of Polymarket ’

TL;DR

I was in the middle of a discussion about RFK’s Trump endorsement when one participant confidently declared that it improved Trump’s odds by +2% because Polymarket moved by +2% from rumor to endorsement. It was fine observation as the event happened relatively quickly and there wasn’t much additional news that could have moved the market. It might have been a fair statement if these were efficient markets.

The problem is that Polymarket is still an inefficient and emerging market that cannot give you good information about <5% moves.

The way efficient markets give you information about <5% moves is that a reasonable number of investors trade on the information. If you believe an endorsement helps Trump’s odds by +10% then you buy with size (and leverage) to earn a decent return. If you believe it hurts Trump by -2% then you sell with size (and leverage). The market is able to take a weighted average of thoughtful predictions to reflect updated information.

The problem is that Polymarket experiences large price impact relative to small deviations from consensus. In the market below, I may see Trump at 50% odds and may believe an RFK endorsement puts him at 55%. The problem is that I can’t realistically put on a $1M trade to earn $100K (assuming that I am right about ~50% to ~55%) because buying $1M of shares would blow out my average price per share to 62% and I would actually lose money if the market readjusted to 55%.

These problems can be solved by better liquidity and the ability to leverage prediction tokens. Improved liquidity would allow me to place a $1M bet on Trump at 50% and exit at 55% without slippage to earn a 10% return. At a first (and slightly arbitrary) approximation, I believe Polymarket is efficient at the price impact level of a $1M bet. The market will be efficient within <5% when a $1M bet moves the market <5%.

The ability to leverage tokens could substantially improve Polymarket’s efficiency. I believe a 10% return is low on this type of investment — you have to be right and take on emerging crypto platform risk for ~10% assuming you are certain about the outcome. If you think you have a ~80% probability of being right and lose 5% in the case you are wrong, then we are talking about a ~7% expected return on what should be a great prediction relative to the market consensus. It’s too low.

The ability to leverage Polymarket shares would allow me to structure a 25% trade by leveraging up my Yes shares 4-to-1 with enough left over to pay the lending platform 5% for giving me the loan. This is a more appropriate return on capital for a researcher that spent energy and effort to find a good trade and took risk on his conclusion.

Traditional markets can debate 20bps moves in interest rates because they have sufficient liquidity and leverage for a whole ecosystem of analysts to earn high returns by being right about small moves. Polymarket can debate ~20% moves in election odds and maybe even ~10% moves but it is not efficient enough to give you information at the <5% level.

The counterargument is that Polymarket has reached sufficient scale to benefit from wisdom of the crowds. The idea is that tens of thousands of small traders (including many thousands of traders from swing states) constantly make small bets on the election using the latest information. These investors have biases and incomplete information, but the sources of bias are uncorrelated and cancel each other out while the market aggregates many disparate sources of information.

The problem with this counterargument is that Polymarket has ~4,000 daily active traders betting on US elections, and these traders cluster in New York City and California and disproportionately follow the same accounts on Twitter. Of these ~4,000 traders, I would anticipate that half of them are trading without doing deep diligence on each investment only to benefit from airdrop farming — i.e. trading to earn Polymarket tokens as Polymarket is likely to reward active traders with an airdrop. Then there is the issue of concentration as 20 traders account for 95% of Polymarket’s volume.

Polymarket is a wonderful business and source of information. I believe it is much better than polls, pundit models, and opinion pieces at forecasting major political events. Polymarket has demonstrated an ability to predict large moves before they are picked up by major media outlets. That doesn’t mean that Polymarket is a surgical prediction tool that can debate <5% moves like the liquid financial markets that many of us are used to.

Disclaimer：

1. This article is reprinted from [@TheiaResearch]. All copyrights belong to the original author [@TheiaResearch]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.