When it comes to the US election the betting markets are wrong

Opinion: The polls in the US presidential election remain agonisingly close. But the betting markets tell a vastly different story. If you listen to the betting markets then Trump has the election sewn up tight.

Here is how the betting markets work. You can buy a share that Trump will win or that Harris will win. If Trump wins then each person holding a Trump share will win $1; if Harris wins then each Harris share will pay $1.

Currently, Trump shares are selling for about $0.60 while Harris shares are selling for $0.40. If Trump wins, then you get $1 per share; a 0.40 profit per share. The easy way to think of this is that the markets are saying that Trump has a 60% chance of winning while Harris is at 40%.

Many believe that these markets are more reliable than the polls because here people are actually putting their money where their mouth is. This is no longer cheap talk.

Markets are very good at aggregating information held by numerous disparate individuals. But I believe the betting markets are overestimating Trump’s chances of victory.

When it comes to markets, there are two things that can happen. One is called the “wisdom of crowds” and the other is “information cascades”.

Here is what the “wisdom of crowds” approach suggests. Suppose you ask many people to guess the number of gumballs in a very large container. It is highly likely that most of the guesses will be wildly off the mark. But if you take the average of all those guesses, you will come surprisingly close to the correct answer. Why? Simply because in making their guesses, people will err on either side; too little or too much. When you average these out the answer comes close to the actual answer.

But two things are critical for this process to work well. First, you must have many people guessing. Second, these people must be making a guess on their own, independently without being influenced by others, particularly those who guessed before. As long as these two conditions are met, wisdom of the crowds will provide a good answer.

When may things not work well? When either or both of the two conditions are violated; the market is “thin” with few participants, and/or people are not making independent decisions but rather basing their decision on what someone else thinks.

Have you ever been to a highly rated movie, show or restaurant and left thinking: “What are people talking about? That wasn’t so great.”

It is highly likely that plenty of others thought the same, but they kept their silence thinking, that if a movie critic, restaurant critic or social media influencer with thousands of followers thinks so, then it must be true; that, “there must be something wrong with me that I did not find it that great”. And therefore, we often go along with the received wisdom even though privately we may feel differently.

We are no longer making an independent decision based on what we think is true, but rather basing our decisions on what someone else thinks is true, especially people who decided before us.

There is an “information cascade” here where someone states something to start with. For instance, Elon Musk says that Trump will win and puts money on it. Then others follow suit and bet on Trump winning, not because each individual is convinced that Trump will win but because Elon Musk thinks so.

If that happens then the ability of that market to predict correctly is compromised.

The other issue with the betting markets is that they may be dominated by a few big players. The New York Times reports that out of about US$100 million in PolyMarket, nearly 30 percent (US$28 million) came from a single bettor betting from four different accounts.

It is possible that this person really thinks that Trump will win. But it is also possible that while others are donating millions to political action committees, this person decided to put the money into a betting market. The aim being to convey the impression that Trump will win. This may help raise donations, create enthusiasm and also turn out the vote.

And it is possible that many others who are betting for Trump are being guided by a large bettor like this. These smaller bettors are betting not because they think Trump will win but because someone else thinks Trump will win; an information cascade leading to a false consensus.

We will find out soon enough, and no, I have not bet any money on this.