Whales stir the waters, Trump predicts soaring market odds, manipulation or misinterpretation?

CN
7 months ago

Trump is also rising in the aggregate polling data, so his actual chances of winning may be increasing.

Written by: Aubrey Strobel

Translated by: BitpushNews

Recently, on the cryptocurrency-based prediction market Polymarket, the former president Donald Trump's winning odds have been boosted by a small group of traders, possibly even a single individual, through substantial betting. This phenomenon has sparked widespread attention and discussion. Supporters of Vice President and Democratic candidate Kamala Harris expressed their concerns on the X platform (formerly Twitter), claiming that the market is being manipulated. The Wall Street Journal speculated that this surge in odds might be an illusion created by four Polymarket accounts that collectively wagered about $30 million in cryptocurrency on Trump's victory.

Indeed, it is possible that someone has placed heavy bets on Trump. However, this does not necessarily imply any wrongdoing.

First, it is evident that Trump's odds are rising not only on Polymarket. Currently, the trading price for Trump's "win" shares on Polymarket is about 59.9 cents, indicating that the market believes he has a 59.9% chance of winning (each share pays $1 if the prediction comes true, otherwise it is worth zero). Predictit—a U.S. platform with strict betting limits (theoretically should not be influenced by Polymarket whales)—gives Trump a winning probability of 56%. Kalshi, a regulated platform in the U.S., gives a probability of 58%. If you think this price action is "manipulation," you might wonder why Kalshi—a platform where manipulative behavior would be reported—has trading conditions consistent with Polymarket.

Online bookmakers also show Trump leading: BetOnline gives 59%, Betfair gives 55%, Bovada gives 60%, and so on. Moreover, Trump is also rising in the aggregate polling data, so his actual chances of winning may be increasing. Trump's rise in the prediction market coincides with Harris's poor performance in a Fox News interview, and new polls show Trump's support is on the rise.

Secondly, a single anonymous entity placing large bets is not, in itself, evidence of "manipulation." Bloomberg columnist Matt Levine explicitly stated: "This does not look like market manipulation: Fredi9999 is not making hasty purchases to push prices up, but is buying cautiously, which seems designed to acquire a large number of Trump contracts with his money." The simplest explanation is that a trader may genuinely believe Trump is undervalued and is willing to place heavy bets on him.

Just because there is a large buyer does not mean manipulation. The entire premise of the market is to compress available information into prices by rewarding those willing to take risks to express their views. The identity of the traders or the distribution of trades is irrelevant; theoretically, everyone has the incentive to extract information from the market when their view of an asset's fair value contradicts the market value. The market does not need to be democratic to be reliable. It only needs the participants who understand the situation best to express their views financially. The concentrated ownership of assets by a single trader does not delegitimize the price in any way. No one questions the price of Apple stock because Warren Buffett's Berkshire Hathaway owns a lot of it.

The market never "gets it wrong." It merely reflects all available information. If you correctly oppose the market view, you can be rewarded by betting on your own beliefs. Due to regulatory and settlement issues, U.S. users cannot use Polymarket, but they have other options. If you believe that a whale on Polymarket has significantly inflated the price of Trump contracts and that they are b, you can simply bet against them by going long on Harris. While this is not without risk—Harris still needs to win for your bet to pay off—if you believe her "true" chances are 55%, then you are buying something worth 55% at today's price of 40%. Even if you may not want to do this, other market participants will. So if the whale on Polymarket is indeed misinformed, now that we know there is a (possibly misinformed) whale, you would expect the odds to decrease as traders incorporate this new information. Unless prediction markets are generally reliable, and the whale has not significantly impacted them.

But let’s assume traders are indeed spending large amounts to "paint the tape," making Trump's victory appear more likely.

First, this may not even be beneficial for Trump. It could lead to complacency among his voters, reducing their turnout on election day (and possibly motivating Harris's supporters to vote). Secondly, that $30 million could be spent on advertising in battleground states, thereby substantially impacting the campaign. Polymarket remains a relatively unknown platform, so any public relations benefit gained from slightly moving the odds is extremely difficult to quantify.

Moreover, as Polymarket pointed out in a detailed blog post on Monday refuting the manipulation allegations, "If any user has reason to believe that trades are being made for reasons other than financial ones, they can easily adjust the odds by excluding the whale's bets." Why pay attention to prediction markets? In terms of elections, markets convey information faster than experts, polls, or media. This trend will continue into the November elections, not only for major presidential contracts but also for state-level contracts. Traders have financial incentives to reveal non-public information, allowing election observers to see the winning probabilities in all 50 states updated in real-time, without waiting for news from Fox or CNN. By leveraging traders' enthusiasm, markets can and do outperform designated experts.

Polymarket has consistently been a more reliable source of information. Even before President Joe Biden's disastrous debate performance, Polymarket traders indicated a significant probability of his withdrawal, which mainstream experts overlooked.

Additionally, contracts like the election winner provide us with potentially more useful information output than mere polling aggregations. National polls showing Harris leading 49% to Trump's 47% do not tell you much about her actual chances of winning: of course, the total electoral votes matter, especially what happens in battleground states. The highly liquid Polymarket election winner contracts compress all this information, providing us with a simple estimate of each candidate's chances of winning.

Remember, a 60% to 40% lead in prediction markets is fundamentally different from a 60% to 40% lead in polls. Prediction markets should be viewed as high-beta derivatives of polls. If Trump leads Harris 51% to 49% in national polls, his actual chances of winning would be 90% or higher. In prediction markets, odds of 40% to 60%, 45% to 55%, or 60% to 40% are all in the realm of coin flips. So Trump's rise on Polymarket does not reflect a massive reassessment of his winning chances. Regardless of whether there are whales on Polymarket, this remains an extremely competitive race.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。

Bitget:注册返10%, 送$100
Ad
Share To
APP

X

Telegram

Facebook

Reddit

CopyLink