What is a prediction market?

A prediction market is where people can put money to speculate about a possible outcome of a future event. If you managed to predict the right outcome, you can reap in some profits. Predict wrongly and you lose money. You can buy contracts and shares of the different outcomes, and trade them in the prediction market as you would with a stock. Here’s a hypothetical example:

There are a couple of players in the market, and each of them can use the prediction market to profit and contribute to the information aggregation.

Suppose Bill is an investor looking to gather market sentiment on Apple products. Bill wants to know if Apple could sell more than 250 million iPhones in 2019 (They sold about 216.76 million phones in 2017). He could create a contract “More than 250 million iPhones sold in 2019”. The contract starts off costing $50. The agreement would be as such

“This contract cost $50. If Apple sells enough, you can claim back $100. If not, you get back $0”

Suppose in the middle of the year, the contract is trading around $36. This signals that the probability of Apple achieving 250 million iPhone sales is about 36%. Overall, people are willing to bet $36 on this outcome, to win $100. From the investor’s perspective, he can take it as a sign, that the market believes that Apple’s sales might not continue to grow, and is a reason to be bearish.

From a perspective of a trader, they can profit if their view on the issue is different. By participating, they will contribute their small piece of information into the price. In comes Steve, who is more optimistic about Apple’s sales. He thinks the probability of hitting 250 million is about 75%. Anything under $75 would be undervalued to Steve, and he would buy these contracts. The price of the contracts would rise, which means the market thinks that the outcome is now more likely to happen.

This way, Steve contributed to the information aggregation. If Apple hits the target, Steve profits $64. If Bill saw a rising probability of Apple hitting the target, he could confidently invest in Apple and earn money as well. Profits everywhere!

Examples of prediction markets

Some of the existing prediction markets are used to predict presidential elections, like the Iowa Electronic Market (IEM) and PredictIt. Or which sports team might win a championship, like TradeSports.

One of the more interesting prediction markets would be the Hollywood Stock Exchange. On HSX.com, traders can buy and sell “stock” of celebrities and movies. The stock prices act as predictions on how much that particular movie would gross at the box office. Stock prices of celebrities (talk about objectification, eh?) would change based on their success, rising if the movies they acted in were a hit, and falling if their movies were a miss.

All these markets serve as a platform for people to place their bet, based on what they think might happen. If I strongly believed the Lakers could win, I could buy Lakers contracts, and I would get a payout if they win.

Suppose I bought the Lakers contract a week before the game. But a night before the game, news of a member from the opponent team getting injured is announced. Now the possibility of the Lakers winning is higher, and the price of the contract has increased.

Just like stocks, I could either sell my contract on the market for more now, or I could hold on till the results of the game is out. I might earn more from making the right prediction. OR, I might be slapping myself because they lost, along with my bet and the profits I could have made.

Prediction Markets vs Stock Markets vs Betting Markets

Could the stock market be considered a prediction market? In some ways, yes. You can trade contracts on a prediction market like stocks. Efficient market hypothesis suggests that the prices in both markets are accurately priced, and reflects all information. Both markets can therefore be used to gather information about whatever is on the market.

However, the distinction is pointed out in “Prediction Markets” by Luckner et al. Once we look into the contracts within each market, we notice a fundamental difference: assets in a stock market are based on an underlying real asset: the ownership of an existing company. In contrast, assets of a prediction market are possibilities, which do not have real value by themselves. Luckner et al writes,

In contrast to prediction markets, however, stock markets are established with the primary purpose of allocating resources, trading risk, and rising capital. Information aggregation is only a pleasant by product of stock markets…

The value in a prediction market is in the aggregation itself. The price reveals how probable the market thinks an event is going to happen. We get a glimpse of the most likely outcome and can make better decision based on those probabilities.

“So what’s the difference between this and a betting market?” Indeed, in markets where we can gamble on which sports team will win the championship, or which horse will win the race, prediction markets seem to be no different.

One could argue that betting markets serve as entertainment, where the reward is not really about the money, but about the thrill of winning a gamble. Prediction markets serve to aggregate and reveal information, and winning money is the incentive — a means in order for people to contribute the best information they have at hand. Betting markets would then be a subset of prediction markets, the “just for fun” aspect of prediction markets.

How accurate are prediction markets?

In March 2000, HSX.com successful forecast all the winners in the six major Oscar categories (Surowiecki). And in 2006, it accurately predicted 32 out of the 39 major category Oscar nominees and seven out of eight top-category winners.

Luckner et al. compared the accuracy of the STOCCER Championship Market, another prediction market, against other prediction methods (i.e. FIFA world rankings, random draws, betting odds from other platforms) and found that STOCCER would be “about as accurate as betting odds and more accurate than the FIFA ranking and random predictor”.

To emphasize how powerful these figures are, these markets were up against experts in the field, people with supposedly more knowledge than most people. When HSX.com forecast all the winners of 2000, they beat reporters from The Wall Street Journal. The reporters spoke to about 350 members in the Academy to find out who they voted for, and managed to predict 5 out of the 6 winners correctly.

Imagine that. Professional reporters. Years of experience and the opportunity of obtaining insider information and yet they still lost to online players trading play money on a website selling Angelina Jolie stocks. It’s like a teacher who peeked a little at the answers on a test, but still lost to the entire class working together.

The power in prediction markets

What gives prediction markets their fearsome ability to give accurate predictions about the outcomes of certain events? The answer lies in how it taps into the power of the collective. In James Surowiecki’s book “The Wisdom of the Crowd”, he describes how the wisdom of the crowd works. As long as a crowd meets certain conditions, it will give rise to a wise crowd. Surowiecki lists out the four conditions, namely:

Diversity of opinion — Every person has to possess some knowledge on the subject at hand, and have their own views about it

In “The Wisdom of the Crowd”, Surowiecki says that you don’t have to be an expert about a topic to make a prediction (in fact, he argues that it might be counterproductive if everyone is an expert). They just have to know some information, and then generate a view.

It is important to emphasize that while the barrier to entry is low, pulling predictions out of thin air will not get us anywhere. A completely uninformed crowd has no way of beating an expert. However, every individual is free to interpret that information available to them however they like, no matter how strange or eccentric.

This diversity means we can see different perspectives on the same manner, something that might not be achieved by a single individual. Differing perspectives mean that whatever destructive errors an individual might make are cancelled out by another person’s input. Given a diverse enough group of people, Surowiecki argues that the errors in each of them will cancel themselves out, leaving only the best information as a result. A prediction market’s price would therefore reflect just that: A buyer of a contract might have some opinion on the matter that would be counteracted by a seller, and vice-versa. A wise crowd eliminates the errors of the individual.

Independence — In order for effective crowd wisdom, people’s opinions should not be determined by the opinions of those around them.

Rightfully so, if we want to have an accurate picture of reality and the future, honesty is key. When people hide or change their opinions in order to fit in with the rest, it results in what is known as “groupthink”, a phenomenon where the group’s collective thinking becomes tunnel-visioned, because a stronger member within the group has forced, consciously or unconsciously, to only one opinion. A prediction market’s result, would be greatly skewed if everyone blindly followed the popular opinion when placing a bet. This reduces its effectiveness in painting an accurate picture.

Decentralization — people are able to specialize and draw on local knowledge

Decentralization has been very popular amongst blockchain and cryptocurrency advocates, but it is not a new phenomenon. In the book “The Starfish and The Spider”, Ori Brafman and Rod A. Beckstrom discussed the powers of a decentralized network, giving the example of Wikipedia, as well as the famous Burning Man festival. They point out a few principles that power decentralized systems, some of which are

An open system does not have central intelligence, and the intelligence is spread throughout the system

Put people into an open system and they’ll automatically want to contribute

The unique point about decentralization is that it fosters responsibility and ownership in every user. There is no Wikipedia police, nor law enforcement at Burning Man. That means every individual in that system has to play that role, to ensure that no one maliciously tempers with articles, to make sure the gift economy of Burning Man runs smoothly. When people are empowered with these roles, they want to contribute to the community. Everyone brings their own unique set of skills and specialization to the table, working hand in hand to solve problems together.

For a wise crowd to work, there is no leader to dictate who is destined to give what opinion. There is only a general direction, a question to be solved, which provides the only common ground for people to work together. In the case of a prediction market, they have to solve or provide possible outcomes of certain events.

Aggregation — some mechanism exists for turning private judgments into a collective decision.

When tapping into the wisdom of the crowd to locate the missing U.S. submarine Scorpion, John Craven used Bayes’ Theorem to create an accurate estimate of the Scorpion’s location. When the Challenger exploded, the stock market managed to pinpoint which was the company responsible within 13 minutes. These are examples of mechanisms that harness the power of a wise crowd.

Everyone’s information needs to find a way to be collated in order to produce a coherent picture. Prediction markets use the trading prices of contracts as the mechanism for collating and aggregating information together.

The future of prediction markets

Someday, when prediction markets are big enough, the possibilities are endless. For now we’ll have to settle for crystalball gazing as the biggest prediction market.

A couple of probable scenarios mentioned by Youtuber Blockwolf — If you were moving to another state for a long time, should you buy/rent? Why not ask the crowd, by creating a contract in the prediction market to know “will property prices in Chicago rise in two years time?”. If the contract is trading for a high price, chances are property prices will rise, and it might be a good decision to buy instead of rent. What if you are a small business and would like to choose Snapchat to advertise your services? Perhaps asking “Will Snapchat have a greater user base in 2019?” might yield the answer.

Some envision decentralized prediction markets as the future market for insurance. Insurance companies would be able to accurately and fairly provide insurance policies based on predictions of when disaster will strike within a particular area.

One could even hedge their own actions. Bought a ticket to the water-park, but afraid that it would rain that day? Bet on it in a prediction market. That way, if it does rain, you might have lost money for your water-park ticket, but you would have minimized that loss with your profit from your bet.

These are grand visions of the future, where nothing is off limits. However, in order for such a vision to work, it cannot be on the scale of prediction markets today. The number of participants would not be enough to answer the few questions above, because only a small percentage of participants would have any information at all about the topic at hand. What would I know about California property prices if I lived in Asia? How many people on the current prediction market platforms have any inkling of such an issue?

This is where a decentralized prediction market platforms might come into play. Augur is such a platform, and it has only just started. In the next article, let’s explore what is Augur, how is it a step up from centralized prediction markets, its pros and cons and an introduction of how to use it.

References

Apple iPhone unit sales worldwide 2007–2018 | Statistic. (n.d.). Retrieved from https://www.statista.com/statistics/263401/global-apple-iphone-sales-since-3rd-quarter-2007/

Augur (REP) Explained | What is Augur | BlockWolf. (2018, May 28). Retrieved from https://www.youtube.com/watch?v=TuEm1srIGDs

Augur | A Decentralized Oracle & Prediction Market Protocol. (n.d.). Retrieved from https://www.augur.net/

Bayes’ theorem. (2018, August 18). Retrieved from https://en.wikipedia.org/wiki/Bayes'_theorem

Brafman, O., & Beckstrom, R. A. (2006). The starfish and the spider: The unstoppable power of leaderless organizaitons. London: Portfolio.

R. (2013, January 20). Burning Man: The Documentary. Retrieved from https://www.youtube.com/watch?v=H0NdVn2O1WY

Henry B. Tippie College of Business. (n.d.). Retrieved from https://iemweb.biz.uiowa.edu/

Luckner, S., Schröder, J., Slamka, C., Franke, M., Geyer-Schulz, A., Skiera, B., . . . Weinhardt, C. (2012). Prediction Markets Fundamentals, Designs, and Applications. Wiesbaden: Gabler Verlag.

PG-13, G. A., PG-13, G. T., PG-13, G. S., R, G. H., N/a, G. A., N/a, G. F., . . . N/a, G. D. (n.d.). Retrieved from https://www.hsx.com/

Prediction market. (2018, August 28). Retrieved from https://en.wikipedia.org/wiki/Prediction_market

Staff, I. (2018, June 29). Efficient Market Hypothesis (EMH). Retrieved from https://www.investopedia.com/terms/e/efficientmarkethypothesis.asp

Surowiecki, J. (2014). The wisdom of crowds why the many are smarter than the few. London: Abacus.

The Prediction Market for Politics. (n.d.). Retrieved from http://www.predictit.org/

The Wisdom of Crowds. (n.d.). Retrieved from http://wisdomofcrowds.blogspot.com/2009/12/stock-market-reaction-to-challenger.html

TradeSports. (2018, August 26). Retrieved from https://en.wikipedia.org/wiki/TradeSports

Shoutout to my friends who helped proofread my article, hope you all enjoyed it!